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
|
|---|---|---|---|---|---|
Bcl-w belongs to the prosurvival group of the Bcl-2 family , while the glutamate receptor δ2 ( Grid2 ) is an excitatory receptor that is specifically expressed in Purkinje cells , and required for Purkinje cell synapse formation . A recently published result as well as our own findings have shown that Bcl-w can physically interact with an autophagy protein , Beclin1 , which in turn has been shown previously to form a protein complex with the intracellular domain of Grid2 and an adaptor protein , nPIST . This suggests that Bcl-w and Grid2 might interact genetically to regulate mitochondria , autophagy , and neuronal function . In this study , we investigated this genetic interaction of Bcl-w and Grid2 through analysis of single and double mutant mice of these two proteins using a combination of histological and behavior tests . It was found that Bcl-w does not control the cell number in mouse brain , but promotes what is likely to be the mitochondrial fission in Purkinje cell dendrites , and is required for synapse formation and motor learning in cerebellum , and that Grid2 has similar phenotypes . Mice carrying the double mutations of these two genes had synergistic effects including extremely long mitochondria in Purkinje cell dendrites , and strongly aberrant Purkinje cell dendrites , spines , and synapses , and severely ataxic behavior . Bcl-w and Grid2 mutations were not found to influence the basal autophagy that is required for Purkinje cell survival , thus resulting in these phenotypes . Our results demonstrate that Bcl-w and Grid2 are two critical proteins acting in distinct pathways to regulate mitochondrial morphogenesis and control Purkinje cell dendrite development and synapse formation . We propose that the mitochondrial fission occurring during neuronal growth might be critically important for dendrite development and synapse formation , and that it can be regulated coordinately by multiple pathways including Bcl-2 and glutamate receptor family members .
Mitochondria have been shown to undergo morphological changes in many neurodegenerative and psychiatric diseases , suggesting their vital role in maintaining the normal function of neuron cells . One of the morphological changes in mitochondria is the length or size , which can be controlled by mitochondrial growth or mitochondrial fission/fusion cycles . Mitochondria are dynamic organelles that can undergo fission , fusion , branching , and change in subcellular distribution [1]–[3] , resulting in the exchange of their genetic materials , alteration of their shape , and increase or decrease of their number [1]–[3] . This dynamic nature of mitochondria is also critically important for energy generation , calcium buffering , and control of apoptosis . Mitochondrial fission and fusion is normally a well-balanced event; when the fission is blocked , the length of mitochondria increases due to ongoing fusion , and mitochondrial fission sites persist as constriction sites due to the slowdown of fission , whiles when the fusion is inhibited , mitochondria usually appear fragmented [3] . Mitochondrial number increases during cellular division , growth , and differentiation via the fission process [4] . However , excessive fission can stimulate apoptosis [5] , and cause neurodegenerative diseases [6] . In cultured healthy neurons , mitochondrial fission and fusion proteins have been shown to regulate the morphology and plasticity of dendritic spines and synapses [7] . In addition , glutamate [8] and synaptic activity [7] modulates the motility and fusion/fission balance of mitochondria and controls mitochondrial distribution in dendrites [7] . Several proteins have been identified in a variety of species to mediate mitochondrial fission or fusion process [2] , [3] , however , little is known about the signaling molecules that activate these processes . Cerebellar Purkinje cells are characterized by large and highly branched dendritic arbors in the brain . Over 90% of Purkinje cell dendritic spines form excitatory synapses with granule cell parallel fiber axons , which relay information from pre-cerebellar nuclei to Purkinje cells . Grid2 is strongly expressed in Purkinje cells [9] , and localizes specifically to Purkinje cell/ parallel fiber synapses [10] , [11] . Analysis of Grid2 knockout mice [12] , and Hotfoot mice carrying spontaneous loss-of-function mutations in Grid2 [13] , [14] has demonstrated that these mice exhibit an impaired function on motor coordination and learning tasks , and have structural and functional defects in Purkinje cell/granule cell parallel fiber synapses and altered long term depression [12] , [15] , [16] . Physiologic studies of Grid2Lc , the Lurcher dominant mutation have established that the Grid2Lc mutation results in inward Ca2+/Na+ current and constitutive activation of the δ2 glutamate receptor , and also that the Grid2Lc receptor has similar channel properties to both NMDA [17] and AMPA receptors [18] , [19] . Activation of Grid2Lc also induces autophagy and degeneration of Purkinje cells . This degeneration might be mediated through interaction of Grid2 with its downstream autophagy protein , Beclin1 [20] . Autophagy is a conserved mechanism for degradation of proteins and other subcellular constituents , and is often involved in cell and tissue remodeling or cell death [21] . Two recent reports demonstrated that Purkinje cells also degenerate without the presence of the basal level of autophagy [22] , [23] . Bcl-2 family members have been most extensively studied in the context of apoptotic cell death [24] . The Bcl-2 family was divided into the pro-survival members that protect cells from being killed , and the pro-death members that kill cells . Bcl-w belongs to the pro-survival group of the Bcl-2 family that includes Bcl-2 , Bcl-XL , A1 , and CED-9 [25] , [26] . These proteins function to protect cells from apoptosis by binding to the outer membrane of mitochondria through their C-terminal hydrophobic domain , thereby preventing the release of several apoptosis proteins from mitochondria into the cytoplasm . They include the caspase regulatory proteins and proteins that lead to DNA fragmentation and chromosome condensation [27] . Bcl-w is widely expressed in a variety of tissues , but predominantly in adult brain and spinal cord [28] . The expression of Bcl-w in brain increases during the postnatal development and is maintained at high levels in the adult brain including cerebellum Purkinje cells , where it is localized to Purkinje cell soma [29] and dendrites ( Lab Vision Corporation ) , whereas Bcl-XL , the only other pro-survival member that is expressed in adult brain had much lower level of expression [29] . Bcl-w−/− mice are smaller during the early postnatal development , but viable and normal in appearance as adults . Both apoptotic and non-apoptotic cell death have been observed in the testes of Bcl-w−/− mice [30] , [31] . A recent report as well as our own findings demonstrated that several other survival members of the Bcl-2 family including Bcl-w could also bind to the autophagy protein , Beclin1 [32]–[34] . Beclin1 has been shown previously to form a protein complex with an adaptor protein , nPIST , and the intracellular domain of Grid2 [20] . Thus , Bcl-w might interact genetically with Grid2 to regulate mitochondrial , autophagy , and neuronal function . In this study , we aim to understand how Bcl-w and Grid2 interact genetically to regulate mitochondria , autophagy , and neuronal function using Bcl-w and Grid2 null mutant mice . We show that the survival member of the Bcl-2 family member , Bcl-w does not control the cell number in brain , but promotes what is likely to be the mitochondrial fission in Purkinje cell dendrites , and is required for the Purkinje cells/parallel fibers synapses and motor learning . We demonstrated that the excitatory receptor Grid2 could regulate mitochondrial length , and the mutation of this protein shares the similar phenotypes in cerebella with the loss-of-function of Bcl-w . Comparative analyses of single and double mutant mice of Bcl-w and Grid2 further indicate that these molecules act synergistically to regulate mitochondrial length and to control the development of Purkinje cell dendrites , dendritic spines , and synapse formation . We further show that no evidence of alteration of autophagy in single and double mutant mice was observed , and the potential upregulation of Beclin1 in Bcl-w−/− mice and overexpression of Beclin1 was not sufficient to activate autophagy . We have thus identified Bcl-w and Grid2 as two critical proteins acting in distinct pathways to control mitochondrial morphogenesis and Purkinje cell development in the mouse cerebellum .
Since Bcl-w binds to Beclin1 , which in turn can form a protein complex with nPIST and the intracellular domain of Grid2 [20] , the possibility arises that Bcl-w may function downstream of Grid2 . We thus examined if Bcl-w−/− [30] , [31] and Grid2ho−4J ( −/− ) mice [14] , [35] , [36] that carry spontaneous null mutation of the Grid2 gene share similar phenotypes . Since Bcl-w binds to the outer membrane of mitochondria to regulate apoptotic activity [26] , we examined both cell numbers ( see below ) and the morphology of mitochondria in Bcl-w−/− and Grid2ho−4J ( −/− ) mice by electron microscopy ( EM ) . Purkinje cells were focused on because Grid2 is only expressed in Purkinje cells [9]–[11] , and Bcl-w also had strong expression in these cells in adult brain [28] , [29] . In these EM micrograph , profiles of mitochondria collected from longitudinal sections of dendritic tracks in electron micrographs appear lengthened in both Bcl-w−/− and Grid2ho−4J ( −/− ) mice ( Figure 1A ) . The lengths of mitochondria were thus measured and quantified . Palay and Chan-Palay [37] have demonstrated using EM method that the mitochondrial lengths in Purkinje cells of wild type mice are ∼0 . 1–0 . 6 μ . In the present study , wild type mice yielded an average value ∼0 . 7 – 0 . 8 μ , with about two third of mitochondria measuring between 0 . 1–0 . 8 μ ( Figure 1B ) . This is similar to the previous electron micrographic estimates [37] . In Bcl-w−/− mice , the average length of mitochondria was increased to ∼1 . 4–1 . 5 μ ( Figure 1A , B ) , similarly to that in Grid2ho−4J ( −/− ) mice , ∼1 . 3–1 . 6 μ . Both numbers are significantly different from that obtained in wild type mice ( Figure 1B ) . In addition , it was notified upon detailed examination of the micrographs that mitochondria in both Bcl-w−/− and Grid2ho−4J ( −/− ) mice often contained points where they became constricted ( Figure 1A ) . This observation suggests that the lengthened mitochondria might be due to the inhibition or slowdown of mitochondrial fission process . To understand if Bcl-w and Grid2 act in the same or separate pathways , we generate double mutant mice of Bcl-w and Grid2 . The reason for this was that if mitochondrial and synaptic phenotypes in Bcl-w−/− Grid2 ho−4J ( −/− ) double null mutant mice was similar to that in either of the single knockouts , then it can be concluded that Grid2 and Bcl-w function in the same pathway; whereas a finding be concluded that the additive or synergistic phenotypes of Bcl-w and Grid2 mutants would suggest that they act instead in distinct pathways . EM analysis of mitochondria in Purkinje cell dendrites of Bcl-w−/−Grid2 ho−4J ( −/− ) mice indicated that the average value ∼1 . 8–1 . 9 μ ( Figure 1A , B ) was significantly longer than those mitochondrial profiles obtained from the single mutant mice ( Figure 1A , B ) . In addition , mitochondrial profiles from Bcl-w−/−Grid2 ho−4J ( −/− ) mice contained frequent thinning and constriction sites ( Figure 1A; blue arrows ) , and in some cases their cristae appeared slightly dilated ( Figure 1A ) , indicating perhaps much slower mitochondrial fission compared to single and wild type mice . The mitochondrial length estimated in Bcl-w−/−Grid2 ho−4J ( −/− ) mice is likely to be much underestimated length because very long mitochondria transit out of plate on very thin EM sections . In order to view the morphology of large mitochondria in Purkinje cells , we thus made thicker , 0 . 5 μ semi-thin plastic sections in distal Purkinje cell dendrites of wild type , single , and double mutant mice ( Figure 1C ) . Intriguingly , many extremely long mitochondria ( often >10 μ ) , which can extend for much of the visible length of Purkinje cell dendrites were frequently found in Bcl-w−/− Grid2 ho−4J ( −/− ) double mutant mice ( Figure 1C ) . In addition , small mitochondria in dendrites that contain extremely long mitochondria seem depleted . However , these were rarely found in single , and not at all in wild type mice . Since the mitochondrial length in double mutant mice seem longer than the sum of that in the single mutant mice in semi-thin sections , this supports strongly that Bcl-w and Grid2 genes interact synergistically rather than additively to control the mitochondrial length , and that the severely increased mitochondrial length in Bcl-w−/− Grid2 ho−4J ( −/− ) double mutant mice might result from significant slowdown of the mitochondrial fission in Purkinje cell dendrites . We next examined if Purkinje cell number was altered in Bcl-w−/− , Grid2ho−4J ( −/− ) and Bcl-w−/−Grid2 ho−4J ( −/− ) mice . The rational for this experiment is that cell death has been observed in testes of Bcl-w knockout mice [30] , [31] and in Purkinje cells of Grid2Lc mutant mice [17] , and that mitochondrial fission can stimulate apoptosis and cell degeneration . In these studies , we found that Bcl-w−/− mice brains appeared grossly normal , and no significant difference in Purkinje cell number ( Figure 2A , B ) , or obvious changes in neuron number in other brain regions compared to wild type mice were observed ( data not shown ) . A similar result was obtained in Grid2ho−4J ( −/− ) mice , agreeing with previous studies of Grid2 mutant mice [38] . Despite the fact that cerebella from the Bcl-w−/− Grid2 ho−4J ( −/− ) animals were obviously smaller and contained an overcrowded Purkinje cell monolayer ( Figure 2A ) , normal Purkinje cell numbers were found in the double mutant mice ( Figure 2B ) . This result suggests that Bcl-w and Grid2 promote perhaps mitochondrial fission in non-degenerating Purkinje cells . Previous EM studies of Grid2 null mutant mice have revealed a large number of naked Purkinje cell dendritic spines , and mismatched connections between the pre- and postsynaptic active zones of Purkinje cell/parallel fiber synapses [12] , [35] . Interestingly , our EM analysis of these synapses in Bcl-w−/− mice also indicated a large number ( ∼44% of total spines ) of naked spines , and several synaptic defects including mismatched connections , shortened active zones , and thickened postsynaptic densities ( Figure 3A; Table 1 ) . Although the average number of naked spines did not change between Grid2ho−4J ( −/− ) ( ∼50% –70% ) and Bcl-w−/−Grid2ho−4J ( −/− ) mutant animals ( ∼65% ) ( Table 1 ) , there were much fewer synapses in the double mutant mice due to the significantly reduced number of dendritic arbors ( see below; Figure 4 ) . Using the accelerating Rotarod behavioral test [12] , [15] , we obtained evidence that the synaptic defects in the Bcl-w−/− mice observed in EM studies may result in deficits in cerebellar motor learning function in these mice ( Figure 3B ) . We found that whereas wild type mice improved their performance with experience on the rotating bar , Bcl-w−/− mice consistently failed to improve in performance throughout the trials ( Figure 3B ) . Interestingly , a similar result has been previously reported in the loss-of-function of Grid2 mice [12] , [35] . To rule out potential neuromuscular abnormalities as the cause of this phenotype , a hanging wire test [39] was performed on Bcl-w−/− mice . No obvious differences in the retention time between Bcl-w−/− and wild type mice were observed ( data not shown ) , suggesting that the lack of motor learning evident in Bcl-w−/− mice in the Rotarod assay is due to defects in cerebellar function . In summary , these results demonstrated that both Bcl-w−/− and Grid2ho−4J ( −/− ) mice had the similar phenotypes including significantly lengthened mitochondria in Purkinje cell dendrites , fewer and malformed Purkinje cell/parallel fiber synapses , and motor learning defects . Visually , Bcl-w−/−Grid2 ho−4J ( −/− ) double mutant mice were immediately distinguishable from wild type and the single mutant animals by the fact that they were smaller in size , moved very little , and when prodded moved with an extremely ataxic gait . These motor difficulties were so severe that these mice could not be properly tested in the Rotarod assay ( data not shown ) . To examine these double mutant mice for histological abnormalities , the Golgi impregnation technique was used to further visualize the architecture of Purkinje cell dendrites and the morphology of dendritic spines ( Figure 4A ) . The Strahler method of ordering ( Figure 4B ) was subsequently applied to obtain quantitative estimates on the impact of these mutations on the complexity of dendritic trees [40] , [41] . In the Strahler method of ordering , each dendritic arbor is assigned as “order” such as primary , secondary , tertiary , etc . Dendritic arbors in each order are then subsequently quantified ( Figure 4B ) . An analysis of the data using the Strahler method demonstrated that the single mutant mice did not have significant differences in each of six Strahler orders compared with that of wild type mice . By contrast in the double knockout mice , Purkinje cell dendritic arbors were reduced significantly that there were four or five Strahler orders compared to six orders in wild type and single knockout mice . Furthermore , the number of branches in each order was also reduced significantly in the double mutant mice in comparison to the single knockout and wild type mice ( Figure 4B; Table 2 ) . Additionally , we also found that ∼45% Purkinje cells analyzed in the double mutant mice contained two dendritic branches rather than the single primary branch that extended from the Purkinje cell soma . This compared to ∼5% Purkinje cells analyzed in wild type and the single mutant animals ( Table 3 ) . Examination of Purkinje cell dendritic spines using Golgi impregnation in 4–6 mice ( Figure 5A–D ) for each genotype revealed that Purkinje cell dendritic spines in wild type , Bcl-w−/− , and Grid2ho−4J ( −/− ) mice all had the characteristic door knob-shaped structure and spacing expected , whiles in Bcl-w−/−Grid ho−4J ( −/− ) mice they were crowded onto dendrites , appeared significantly shorter , and often branched or lacked a clearly distinguishable spine head or neck ( Figure 5A , B ) . Similar spine defects were found in EM studies in an additional three double mutant mice . To quantify the difference in spine length between single and double mutant mice , we measured spine profiles that have their necks connected to dendrites ( Figure 5C ) . The wild type Purkinje cell spine lengths determined by this method ( 0 . 86±0 . 33 μ; Figure 5D ) were agreed very well with previous measurements of mouse Purkinje cell dendritic spine length ( 0 . 87±0 . 21 μ ) obtained using confocal microscopy of Lucifer yellow injected Purkinje cells [42] . In this study , wild type and single mutant mice had similar length of spines , although the spine length decreased significantly by ∼25% in Bcl-w−/−Grid2 ho−4J ( −/− ) Purkinje cells ( Figure 5C , D ) . In summary , these results demonstrated that normal Purkinje cell dendrite development and synapse formation require both Bcl-w and Grid2 . Since both dendritic arbor number and spine length in double mutant mice are much more severely affected than the sum of these defects in single mutant mice , we conclude that the interaction between Bcl-w and Grid2 genes in regulation of dendrite and spine development is synergistic . Two recent reports demonstrated that a basal level of autophagy was required for preventing the accumulation of protein aggregates and inclusion bodies and the survival of Purkinje cells [22] , [23] . Since Bcl-w can interact physically with Beclin1 and inhibit starvation-induced autophagy ( unpublished results ) , and the dominant Grid2Lc mutation can induce autophagy in Purkinje cells , it is possible that Bcl-w−/− and Grid2ho−4J ( −/− ) mutations might potentially alter autophagy in Purkinje cells and result in observed phenotypes in these cells . We thus examined anatomic evidences of autophay in Purkinje cell EM sections of single and double mutant mice . It has been shown previously that morphologic evidence for the activation of autophagy indicated by the presence of autophagosomes was readily apparent in Grid2Lc Purkinje cells [20] . In contrast , inhibition of basal autophagy in Purkinje cells can result in the accumulation of inclusion bodies or protein aggregates [22] , [23] . However , in a careful examination of the Purkinje cell cytoplasm in wild type and single and double mutant animals in electron micrograph , no morphological evidence indicative of alterations in autophagy was observed ( data not shown ) , suggesting that autophagy is unlikely to be the cause of Purkinje cell phenotypes in mutant mice .
During mitochondrial fission process , the mitochondrial fission protein complexes localize on the fission sites , which appear as constriction sites during ongoing mitochondrial proliferation [2] , [3] . When the mitochondrial fission is blocked , the constriction sites persist and can be identified easily under the electron microscope [43] . Both the growth of mitochondria and the mitochondrial fission/fusion processes determine the final size or length of the mitochondria in cells . However , it should be notified that mitochondrial growth alone does not generate constriction sites . The frequently observed constriction sites in mitochondria of double mutant mice strongly support that the lengthened mitochondria in single and double mutant mice are due to the slowdown of mitochondrial fission process . Since small mitochondria are seemingly depleted in dendrites that contain extremely long mitochondria in the semi-thin section of double mutant mice , this also supports that slow mitochondrial fission led to the decreased number of mitochondria . However , we cannot rule out that these mitochondrial phenotypes were due to enhanced fusion process . The survival members of the Bcl-2 family have not been previously reported to regulate mitochondrial length or mitochondrial fission/ fusion in mammalian cells . However , the pro-death members of Bcl-2 family , Bax and Bak have been demonstrated to regulate mitochondrial fission and fusion in both apoptotic and living healthy cells [43]–[45] . For example , in C . elegans , overexpression of EGL-1 can induce mitochondrial fission and apoptosis [46] . In mammals , Bax or/and Bak promote mitochondrial fission in apoptotic cells through regulating mitochondrial fission proteins directly [43] or indirectly [47] . In these cells , Bax could also inhibit mitochondrial fusion [44] . However , in living cells Bax and Bak act oppositely as they function to promote mitochondrial fusion [45] . We show in this study that the survival member , Bcl-w has similar characteristics . Thus , Bcl-w promotes mitochondrial fission in Purkinje cells , whiles in testis it protects cells that should normally die in apoptosis . The excitatory receptors have not been previously reported to regulate mitochondrial length or dynamics . Previous studies though have demonstrated in neuronal culture system that synaptic activity can stimulate mitochondrial fission and clustering to the dendritic spines [7] . In vivo results for this regulation are lacking , however . Our results implicate that the excitatory receptor Grid2 regulates mitochondrial morphology in addition to its previously found regulation of channel activity and other functions . The in vivo analysis of mitochondrial localization in Purkinje cells in the current study [37] indicates that mitochondria are normally present in dendrites , but rarely inside dendritic spines . Thus , the actions of Grid2 from synapses on mitochondria may be indirect because Grid2 is localized in Purkinje cell/parallel fiber synapses . The synergistic effect in Bcl-w and Grid2 double mutant mice on mitochondrial length rules out the possibility that Grid2 promotes mitochondrial fission mainly through regulating Bcl-w , and suggests that other pathways might be responsible for Grid2 in regulation of mitochondrial length or fission/fusion . Our studies in the dominant Grid2Lc mutation demonstrated the extensive mitochondrial fragmentation in cytoplasm of Purkinje cell of Lurcher mice during the postnatal development ( unpublished results ) . This suggests that Grid2 might control mitochondrial length through the mitochondrial fission process by regulating calcium influx . Indeed , calcium has been shown in several studies to stimulate mitochondrial fission by regulating the activities of dynamin and the dynamin-like large GTPase , Drp-1 [6] . Thus , Grid2 is likely to function to promote mitochondrial fission through its channel activity . In this study , we observed fewer Purkinje cells/parallel fiber synapses , an increased ratio of naked spines , and motor learning defect in the double mutant mice . This may be correlated with slowdown of mitochondrial fission in Bcl-w−/− and Grid2 ho−4J ( −/− ) mice , as suggested by our studies . A more direct correlation between mitochondrial fission or fusion and the number of spines and synapses has been demonstrated in primary neuronal culture; overexpression of mitochondrial fission protein Drp-1 in these cells resulted in increased number of mitochondria , correlated with increased number of spine and synapse . In contrast , the expression of mitochondrial fusion protein , OPA1 or dominant negative version of Drp-1 has been reported to lead to fewer numbers of spines and synapses [7] . These results implicate that mitochondrial fission in healthy cells might serve as a means to increase the number of mitochondria to meet energy demands during neuronal growth , or neuronal plasticity , and is likely to be different in mechanism from the excessive mitochondrial fission observed during apoptosis and neurodegeneration . Indeed , the mitochondrial fission during apoptosis can result in the loss of mitochondrial DNAs and lower the function of mitochondria [6] . However , in living healthy human cells , Benard et al . demonstrated recently that when mitochondrial fragmentation was inhibited , a strong inhibition of mitochondrial energy production was observed [48] . Bcl-w is localized in Purkinje cell dendrites and acts on mitochondria , the synaptic defects in Bcl-w−/− mice are thus likely to be the consequence , not the cause of the mitochondrial morphogenesis defects . Since mitochondrial fission resulted from the neuronal excitation is linked to the danger of degeneration , it is intriguing that the survival member , Bcl-w could promote mitochondrial fission , and has protective function to cells as well . In summary , the results in the current study suggest that the Bcl-2 family member , Bcl-w , and the excitatory receptor Grid2 can regulate the mitochondrial fission and thus mitochondrial length in dendrites . Altered mitochondrial length in mutant mice of these genes in turn results in abnormalities in synapse formation in the mice . The Bcl-2 family has not been shown previously to regulate neuronal dendrite development; its effect on neuronal growth has been only associated with cell death . In comparison , the NMDA receptor , one of the glutamate receptor family members has been demonstrated to regulate the activity-dependent dendrite development [49] . However , it is not known if this function of the NMDA receptor has anything to do with mitochondrial morphogenesis or Bcl-2 family members . In this study , we demonstrate that the normal development of Purkinje cell dendrite , dendritic spine , and synapse formation requires both Bcl-w and Grid2 , and their regulation of mitochondrial morphogenesis . Mitochondrial proliferation is a biological process that is associated with cellular division and growth . It takes normally three weeks for Purkinje cells to grow from cell bodies into fully-grown trees with extensive synaptic connections [50] . During this period of time , mitochondrial number also increases significantly . This mitochondrial growing process cannot be exhibited well in cultured Purkinje cells that contain very short and little branched dendritic arbors , unfortunately [51] . The significantly inhibition of mitochondrial fission can thus result in decreased small mitochondrial number and the large size of fused mitochondria that result in reduced mitochondrial motility in Purkinje cell dendrites . These could place an intrinsic limitation on the local energy production and calcium buffering in dendrites , resulting in a failure of perhaps neuronal development and function such as the dynamic growth and branching of dendrites , the development and plasticity of dendritic spines and synapses , channel activities , and the formation of the postsynaptic density , thus leading to the severe morphological defects observed in the double mutant Purkinje cells . The abundance of mitochondrial fission during the Purkinje cell growth is also balanced or controlled by mitochondrial fusion . A recent paper demonstrated that the absence of mitochondrial fusion protein Mfn2 during Purkinje cell development resulted in excessive mitochondrial fragmentation and Purkinje cell degeneration , suggesting that mitochondrial fusion is required to prevent cells from degeneration [51] . Similarly , the reason that we did not observe any Purkinje cell death in Bcl-w−/−Grid2 ho−4J ( −/− ) mice in spite of the extensive loss of dendrites , spines , and synapses is likely due to the protective effect by extensively fused mitochondria in these cells . Mitochondrial fusion has also been shown in cultured cells to protect cells from cell death [52] , [53] . The early developmental defects in Purkinje cell primary branches in Bcl-w−/−Grid2 ho−4J ( −/− ) mice indicate that dendritic defects , at least initially , are caused intrinsically , not due to the granule cell parallel fiber innervations because these innervations occur later than the emergence of Purkinje cell primary branches [50] . It has been hypothesized that mitochondrial respiration and metabolism may be spatially and temporally regulated by mitochondrial morphology and location that can be integrated to multiple pathways of cellular function [54] . The regulation of mitochondrial length that can result from mitochondrial fission or fusion thus might participate in other pathways that control dendrite and spine morphology and synapse formation , such as development , diseases , and in response to many extrinsic factors such as neuronal activity , hormones [55] , [56] , and chronic stress [57] . Since we have only observed the mitochondrial morphology changes in fixed tissues , we hope that we can also demonstrate that Bcl-w and Grid2 can affect the mitochondrial fission or fusion in a real time system . This system can also be used to understand the mechanism for Bcl-w and Grid2 and their family members to regulate mitochondrial morphology or the mitochondrial fission and fusion cycle . The studies on mitochondrial fission or fusion will yield important knowledge for our understanding of the development and function of central nervous system .
Bcl-w knockout mice were obtained from Dr . Grant Macgregor’s lab from Emory University ( currently at University of California , Irvine ) . The generation and typing of these mice were described previously [30] . DBA/2J-Grid2ho−4J ( −/− ) mice were purchased from Jackson lab . These mice carry spontaneous null mutation of the Grid2 gene with exons 5–8 deleted , resulting in a 170 amino acid loss in the N-terminal LIVBP-like domain [14] , [35] , [36] . To obtain Bcl-w−/−Grid2ho−4J ( −/− ) double knockout mice , the male Grid2ho−4J ( −/− ) mice were crossed into the female Bcl-w−/− mice to obtain Bcl-w+/−Grid2ho−4J ( +/− ) mice in F1 generation . Both male and female Bcl-w+/−Grid2ho−4J ( +/− ) mice from F1 generation were selected and crossed with each other to obtain F2 generation mice . Pups in F2 generation demonstrating the “hotfoot” ataxic phenotypes were identified as homozygous Grid2 . Molecular genotyping [30] was applied to distinguish Grid2ho−4J ( −/− ) , Bcl-w+/−Grid2ho−4J ( −/− ) , and Bcl-w−/−Grid2ho−4J ( −/− ) animals . Bcl-w−/−Grid2ho−4J ( +/− ) mice do not have an obvious ataxia phenotype because mice obtained from the cross using Bcl-w+/−Grid2ho−4J ( +/− ) male and Bcl-w−/− female mice did not show the obvious ataxic phenotypes . Mice were perfused with 2 . 5% glutaraldehyde , and cerebella were sliced sagitally , and each slice was then diced into pieces containing 2–3 folia . The tissue pieces were post-fixed in 1% osmium , treated with 0 . 5% aqueous uranyl acetate , and then dehydrated through graded alcohol ( 70 , 95 , and 100% ) . After the treatment with propylene oxide , tissue pieces were embedded in a manner allowing sectioning in the sagital plane in Ducupan ( Fluka ) . The blocks were cured in a 60°C degree oven for 2–3 days . Blocks were cut with a glass knife to get Semi-thin sections of 0 . 5 micron . The sections were then stained with 0 . 25% toluidine blue in 1% sodium borate , and evaluated at the light microscope ( LM ) level to select the tissue orientation of sagital and longitudinal sections . Photographs of mitochondria were taken in the molecular layer approximately 1/3 of the distance between the pia and Purkinje cell monolayer using a 100X oil lens in a Zeiss Axioplan light microscope and MetaVue acquisition software ( Universal Imaging ) . Four wild type , four Bcl-w−/− , three Grid2ho−4J ( −/− ) , and three Bcl-w−/−Grid2ho−4J ( −/− ) were examined . To obtain the EM pictures , the sagital and longitudinal block faces were trimmed , and ultra-thin silver sections were cut with a Reichert-Jung Ultracut E ultramicrotome with a Dupont diamond knife , and collected on copper grids , stained with saturated aqueous uranyl acetate , and lead citrate before examination in Jeol 100 cx electron microscope operated at 80 Kv . EM photographs were taken on dendritic tracks randomly in the molecular layer approximately 1/3 of the distance between the pia and Purkinje cell monolayer in sections collected from several different blocks . Mitochondria profiles were traced and measured from the longitudinal dendritic tracks of each set of photographs . For morphometric analysis of synapses , only one section was collected on each grid . After establishing the orientation , locating the pia and Purkinje cell layer , images were recorded from regions 1/3 down from the pia at primary magnification of 6 , 600X x . The print magnification was 16 , 500X . For locating proximal spines on the Purkinje cell primary branches , the block face was reduced allowing more sections to be collected per grid . The male wild type mice were crossed to the female Bcl-w−/− mice to obtain Bcl-w+/− mice , which were subsequently inbred with each other to obtain littermates of wild type , Bcl-w+/− , and Bcl-w−/− mice . Six to eight weeks old of these littermates were tested on accelerating Rotarod with 0 . 1 round/second as starting speed and 0 . 4 round/second2 as accelerating speed . Retention time were begin to be recorded when the mice were placed on the rotating bar and acceleration was applied , and stopped when they failed to run on the rotating bar . Each mouse was given three trials per day for five constitutive days . The hanging wire test was performed by placing mouse on the top of a wire cage lid , and after mouse grip the wires , lid was turned upside down [39] . The retention time for the mouse to hold wire was recorded . Four Bcl-w−/− and wild type mice were tested , respectively . Mice were intracardially perfused with 4% paraformaldehyde in PBS , and their brains were subsequently dissected and post-fixed overnight . Brains were then dehydrated with 70 , 95 , and 100% ethanol , and treated in organic solvent , butanol for three days before replacing it with paraffin . 10 μ sections were obtained from paraffin-embedded brain . These sections were then treated with xylene to remove wax and rehydrated before staining them in Cresyl violet and being mounted on slides . Sections obtained in the region close to midline were selected and counted for Purkinje cells on all folia . Golgi stain of mouse brain was obtained using FD Rapid GolgiStain Kit ( FD NeuroTechnologies , Inc . ) according to the manufacture’s instruction . The images of Purkinje cells were collected using the DIC microscope ( Zeiss ) with 10X object lens ( Figure 4A , upper panel ) and with 5X lens ( Figure 4A , lower panel ) . To analyze Purkinje dendrite branches using the Strahler method of ordering , a Z-stack of Purkinje cell images was collected using MetaVue acquisition software ( Universal Imaging ) , and 20X water lens , and used to quantify Purkinje cell dendritic branches . The spines and dendrites were photographed from combined images from Z-stack ( Figure 5A ) . The spines on dendritic branches of the first Strahler order ( Figure 5B ) were photographed using 100X oil lens . The number of mice examined was indicated in Table 1 .
|
A neuron cell is composed of cell body , axons , and dendrites . Dendritic spines on dendrites form synapses with axons of other neurons , establishing communication between neuron cells . Dendrite development and synapse formation are therefore important for neuronal function . Although many genes have been previously identified as affecting the development of dendrites and synapses , the apoptosis Bcl-2 family members have not yet been shown to regulate these processes . In this study , a Bcl-2 family survival member , Bcl-w , was found not to affect cell death , but to be required for synapse formation and motor learning in mouse cerebellum . Bcl-w also appears to control dendrite development as double null mutant mice of Bcl-w and the glutamate receptor δ2 ( Grid2 ) have severe defects in Purkinje cell dendrites , spines , and synapses . In addition , Bcl-w and Grid2 act synergistically to promote what is likely to be mitochondrial fission in Purkinje cells . Neither the survival members of the Bcl-2 family nor the excitatory receptors have been demonstrated previously to regulate mitochondrial morphogenesis in brain . We conclude that neuronal dendrite development and synapse formation require perhaps mitochondrial fission that can be controlled by two critical pathways including Bcl-w and Grid2 .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"neuroscience/behavioral",
"neuroscience",
"neuroscience/motor",
"systems",
"neuroscience/neurodevelopment",
"neuroscience/neuronal",
"signaling",
"mechanisms",
"neuroscience"
] |
2008
|
Mitochondrial Morphogenesis, Dendrite Development, and Synapse Formation in Cerebellum Require both Bcl-w and the Glutamate Receptor δ2
|
Central regulators of cell fate , or selector genes , establish the identity of cells by direct regulation of large cohorts of genes . In Caenorhabditis elegans , foregut ( or pharynx ) identity relies on the FoxA transcription factor PHA-4 , which activates different sets of target genes at various times and in diverse cellular environments . An outstanding question is how PHA-4 distinguishes between target genes for appropriate transcriptional control . We have used the Nuclear Spot Assay and GFP reporters to examine PHA-4 interactions with target promoters in living embryos and with single cell resolution . While PHA-4 was found throughout the digestive tract , binding and activation of pharyngeally expressed promoters was restricted to a subset of pharyngeal cells and excluded from the intestine . An RNAi screen of candidate nuclear factors identified emerin ( emr-1 ) as a negative regulator of PHA-4 binding within the pharynx , but emr-1 did not modulate PHA-4 binding in the intestine . Upon promoter association , PHA-4 induced large-scale chromatin de-compaction , which , we hypothesize , may facilitate promoter access and productive transcription . Our results reveal two tiers of PHA-4 regulation . PHA-4 binding is prohibited in intestinal cells , preventing target gene expression in that organ . PHA-4 binding within the pharynx is limited by the nuclear lamina component EMR-1/emerin . The data suggest that association of PHA-4 with its targets is a regulated step that contributes to promoter selectivity during organ formation . We speculate that global re-organization of chromatin architecture upon PHA-4 binding promotes competence of pharyngeal gene transcription and , by extension , foregut development .
Selector genes govern the fates of groups of cells related to each other by virtue of their cell type , position or affiliation to an organ [1] . Genomic methods have revealed that selector genes directly control hundreds , even thousands , of target genes , which define the characteristics of a particular cell type [2]–[6] . For example , the mesodermal factor Twist regulates genes that control mesodermal behaviors including gastrulation , migration and proliferation [7] . The myogenic regulatory factor MyoD directly activates skeletal muscle genes during both early cell-fate specification and later differentiation [4] , [8] . The global regulatory strategy of selector genes raises the question of how targets of broadly active selector genes are expressed selectively at the appropriate times and places . The selector gene pha-4/FoxA plays a broad role in the development and physiology of the C . elegans digestive tract . PHA-4 establishes the diverse cell types of the C . elegans pharynx during early embryogenesis , and drives differentiation and morphogenesis at later stages [9]–[12] . After birth , PHA-4 is required for growth and gonadogenesis in larvae [2] , [13]–[15] and promotes longevity in adults [16] , [17] . The targets of PHA-4 are likely distinct in different tissues and at different developmental stages . For example , numerous PHA-4 target genes have been identified within the pharynx , but most of these are not active in the intestine or gonad [2] , [11] , [18] . Recent chromatin immunoprecipitation data with tagged PHA-4 suggest different genes are bound by PHA-4 at different developmental stages [19] . How is appropriate regulation of PHA-4 target genes achieved ? One mechanism is combinatorial control by PHA-4 with other transcription factors . A single PHA-4 binding site is not sufficient for transcriptional activation , and most foregut promoters carry four or more cis-regulatory elements that contribute towards appropriate spatial and temporal expression [13] , [18] , [20]–[25] . In addition , DNA binding affinity of PHA-4 for target genes modulates the timing of activation [2] , [18] . High affinity sites promote earlier transcriptional onset compared to lower affinity sites , within the context of the intact cis regulatory region [2] . These studies suggest that binding affinity , feed-forward loops , positive feedback and combinatorial control , are necessary to achieve accurate temporal gene expression . However , it is still largely unknown how spatial regulation is accomplished . For example , why are pharyngeal genes active in the pharynx but not in the intestine , despite the widespread expression of PHA-4 in both organs ? Studies have implicated the nuclear periphery for modulation of gene transcription . Active and inducible genes are recruited to nuclear pores [26]–[30] . Conversely , nuclear lamins and their associated proteins have been associated with transcriptional repression and chromatin organization [31]–[36] . Inactive genes are often positioned at the nuclear lamina [37] , and tethering of genes to the nuclear lamina can reduce expression levels [38] , [39] . This effect is not comprehensive , however , as some peripherally-located genes are active [38]–[41] . These results indicate that the nuclear lamina is transcriptionally competent , and raise the question of the nature and degree of lamina-mediated repression . The nuclear lamina of C . elegans is composed of a single B-family lamin ( lmn-1; [34] , [42] , three associated LEM proteins [43] and additional factors [44] . Loss of LMN-1 leads to embryonic arrest by the 300-cell stage , with chromosome bridges between sister cells [34] . Inactivation of the LEM protein emr-1/Ce-emerin has no obvious phenotype on its own and produces viable animals , but inactivation of both emr-1 and a second LEM protein man-1/Ce-MAN1 , causes lethality at around the 100 cell stage with phenotypes similar to those of lmn-1 [43] , [45] . Barrier to autointegration factor BAF-1 is a fourth lamina protein required for chromosome segregation and integrity of the lamina [46] , [47] . BAF-1 associates with cis-regulatory sites within the promoters of eff-1 and aff-1 , and is required to repress eff-1 expression in epidermal seam cells [31] . These data implicate the C . elegans nuclear lamina for transcriptional repression , but the mechanism is unknown . In this study , we probe the role of PHA-4 for pharyngeal gene activation , using artificial chromosomes to monitor PHA-4 binding and activity in living embryos [48]–[52] . We find that PHA-4 associates with its targets long before their activation . This association is restricted to a subset of pharyngeal cells , despite the ubiquitous expression of PHA-4 throughout the digestive tract , and is modulated by the nuclear lamina protein EMR-1/Emerin . Binding of PHA-4 leads to extensive chromatin decompaction and repositioning , in a process that precedes transcription . Previous studies implicated mammalian FoxA factors for local opening of chromatin and inhibition of linker histones [53] . Our data suggest that , in addition to local alterations , FoxA factors can induce large-scale changes in chromatin architecture , which may contribute to the long-range effects of FoxA proteins on transcription and recombination [54] , [55] . These studies provide a framework for understanding the cell-type biases of selector genes for their targets .
Our goal was to explore PHA-4 association with its target genes in living embryos . We chose to analyze myo-2 , which is a well-characterized gene expressed exclusively in pharyngeal muscles [56] , [57] , and pax-1 , which we show below is a PHA-4 target expressed in the pharyngeal marginal cells and some other pharyngeal cell types . To initiate the study , we characterized pax-1 cis-regulatory sites for pharyngeal expression . To analyze pax-1 , we constructed two GFP reporters: a translational fusion within the second exon of pax-1 ( PAX-1::GFP; Table S1A ) and a transcriptional fusion between GFP and the pax-1 translation initiation site ( pax-1::GFP ) . These constructs revealed that pax-1 was expressed in 14 pharyngeal cells , which included nine marginal cells , the e2 epithelial cells and the pm8 muscle , based on morphology , position and co-staining for marginal cell filaments ( Figure 1B , Figure S1 ) . We focus on the marginal cells here . Expression of pax-1::GFP in marginal cells was first detectable in two rows of pharyngeal nuclei shortly after embryonic cell division ceased , at the late-bean to early-comma stages of development ( ) . Expression gradually faded during later embryogenesis and was undetectable in larvae or adult worms . Examination of the pax-1 promoter revealed a consensus PHA-4 binding site between −92 and −98 base pairs ( bp ) upstream of the transcriptional start site ( Figure 1C ) . Gaudet et al . previously showed that three copies of this site were sufficient to activate expression of a heterologous promoter within pharyngeal cells [18] . Conversely , we found that deletion ( ) or mutation ( MutP Figure 1C ) of the predicted PHA-4-binding site eliminated pax-1::GFP expression in 17/19 transgenic lines ( Figure 1C ) . We speculate that the 2/19 lines with residual pax-1::GFP expression in pharyngeal cells may be activated by cryptic enhancers originating from nearby sequences in the array . Interestingly , in addition to loss of pharyngeal expression , the mutant reporters exhibited significant ectopic GFP in the epidermis ( Figure 1C and Figure S3 ) . Together , these results suggest the PHA-4 binding site is required to activate expression in the pharynx and repress expression in epidermal cells . To identify additional cis-regulatory sites within the pax-1 promoter , we performed linker-scanning mutational analysis beginning −115 bp upstream of the pax-1 transcriptional start site ( Figure 1C ) . This survey revealed a second activation site within Delta16 , which we will refer to as mutA: replacement of 10 bp from mutA abolished all GFP reporter expression ( TTGAGATTAA; Figure 1D ) . Scanning mutagenesis also uncovered two negative regulatory regions . First , mutations in either Delta14 or Delta18 generated a high proportion of transgenic lines that expressed pax-1::GFP in additional pharyngeal cells , to approximately 20 cells ( Figure S4A ) . Second , mutations in Delta20 , and to a lesser degree Delta22 , lead to GFP+ cells outside of the pharynx ( Figure S4B ) . In sum , mutational analysis revealed both positive and negative cis-regulatory sites that cooperated with PHA-4 to activate pax-1 within pharyngeal marginal cells . A direct repeat ( TTGAGA ) lies within Delta14 and Delta16 , and an inverted repeat ( AAGCTCT ) lies within Delta14 and Delta18 , suggesting one or both of these may be recognition sites for transcription factors ( Figure 1D ) . These cis-regulatory sites provided a means to examine the role of PHA-4 for pharyngeal gene activation , described below . The mutational analysis suggested that both myo-2 and pax-1 were direct PHA-4 target genes . To test this idea further , we used the Nuclear Spot Assay ( NSA ) to examine association of PHA-4 with pharyngeal promoters in vivo . The NSA allowed us to track PHA-4 binding to promoters in living embryos , with precise spatial and temporal resolution . For this assay , we constructed a transgene array or “pseudo-chromosome” that carried multiple copies of a target promoter and the Lac operator [48]–[52] , [58] . A co-selectable marker ( to identify transgenic animals ) and herring sperm genomic DNA ( to provide sequence complexity without added C . elegans' sequences [59] ) were also included . The pseudo-chromosome arrays carried fusions of CFP::LacI and PHA-4::YFP; CFP::LacI bound to LacO sequences on the arrays and revealed their position and morphology in the nucleus . PHA-4::YFP bound to its promoter appeared as a dense magenta “dot” that colocalized with CFP::LacI . Diffuse PHA-4::YFP in the background indicated binding of PHA-4::YFP to genomic loci . It was previously shown that the NSA method accurately reflected transcriptional regulation , as detected by other methods such as chromatin immunoprecipitation [50] , [52] . C . elegans arrays are relatively stable through mitosis and meiosis , and are incorporated into chromatin [59] , [60] . However , we recognize that pseudo-chromosomes are not replicas of C . elegans chromosomes , and they likely differ from endogenous chromosomes in some regards . We observed multiple pharyngeal cells with PHA-4::YFP enriched on pseudo-chromosome arrays , supporting the notion that myo-2 and pax-1 are direct PHA-4 targets ( Figure 2 and an additional third target C44H4 . 1: Figure S5A ) . The association was detected by the 8E ( endodermal ) stage , which was the earliest stage we could visualize PHA-4::YFP ( ∼100 cells ) . The proportion of embryos with associated PHA-4::YFP remained relatively constant until the two-fold ( pax-1 ) or three-fold ( myo-2 ) stages . The robust association of PHA-4 to its target promoters required a consensus PHA-4 binding site since pseudo-chromosome arrays that carried a promoter with mutated PHA-4 binding sites [2] failed to recruit PHA-4::YFP ( Figure 2 ) . These data reveal that PHA-4 bound target promoters long before they were transcriptionally active , indicating that PHA-4 occupancy did not correlate with transcriptional activity per se , but rather with transcriptional potential . Similarly , vertebrate FoxA2 , which is orthologous to PHA-4 , binds the albumin promoter in mouse endodermal cells long before the gene is active [61] . What are the repercussions of PHA-4 association to target genes ? Previous genetic studies suggested that PHA-4 and its orthologues influence the chromatin environment [11] , [53] , [62]–[64] . For example , PHA-4 recruits the histone variant HTA . Z/HTZ-1 to a subset of pharyngeal promoters , including that of myo-2 [63] , and it interacts genetically with predicted chromatin regulators [11] , [63] . Vertebrate orthologues of PHA-4 associate with chromatin and can block compaction by H1 histones [53] , [54] , [65] . These observations prompted us to examine the chromatin morphology of the pseudo-chromosome arrays . We observed progressive decompaction of pseudo-chromosomes bearing wild-type myo-2 or pax-1 promoters , detectable as a large , diffuse dot ( Figure 3A ) . We quantified the changes by measuring the areas of individual pseudo-chromosomes and analyzing the areas with Cox regression models ( Materials and Methods ) . This analysis revealed that both the number of decompacted pseudo-chromosomes and the degree of decompaction increased over time ( Figure 3B , Table S2 , Table S3 ) . This effect was observed within pharyngeal cells , the eventual site of myo-2 and pax-1 expression , but not in non-pharyngeal cells ( Figure 3 ) . Cumulative areas were larger in the pharynx compared to “outside” the pharynx as early as the pre-bean stage for myo-2 ( p = 7×10−11 ) and the bean stage for pax-1 ( p = 0 . 007 ) . For myo-2 , many pseudo-chromosomes became decompacted prior to transcription at the 2-fold stage and remained decompacted . For pax-1 , decompaction began at the comma stage and was maximal at the 1 . 5 and 2-fold stages , when pax-1 is transcribed ( Figure 3B ) . In sum , pax-1 and myo-2 pseudo-chromosomes underwent decompaction preceding and during transcription within the pharynx . In contrast , pseudo-chromosomes in which PHA-4 binding sites had been mutated behaved similarly in pharyngeal and non-pharyngeal cells , with little increase in size over time ( Figure 4 ) . These observations indicate that PHA-4 is required for large-scale decompaction of chromatin in extragenic arrays . We considered three spurious reasons for changing pseudo-chromosome areas , independent of PHA-4 . First , we examined whether array sizes were a consequence of expanding nuclear size . However , nuclear size remained relatively constant at the stages assayed in this study , and no normalization to nuclear size was necessary ( Figure S5D ) . Second , we tested whether decompaction reflected an artificial interaction between LacI and PHA-4 . However , PHA-4 binding and consequent decompaction of pseudo-chromosomes was observed in transgenic lines lacking LacI protein ( Figure S5B ) . Third , we wondered if 3D volumetric measurements would be more accurate than the 2D area measurements used here . 3D analysis was subject to photobleaching of the YFP signal while collecting Z-stacks , which hindered 3D reconstruction . A comparison of area versus volume measurements in embryos with minimal photobleaching revealed a similar trend in array expansion ( Figure S7 ) . These controls suggest that array decompaction reflects PHA-4 interactions with target chromatin . PHA-4 is a critical regulator of pharyngeal gene transcription , and transcription is often associated with chromatin decondensation [66]–[69] . We therefore tested whether decompaction of pseudo-chromosome arrays in pharyngeal cells reflected PHA-4 binding or transcriptional activity . When the PHA-4 binding site is mutated in the myo-2 promoter , myo-2 is still transcribed , but at a later developmental time [2] . We observed little pseudo-chromosome decompaction for arrays bearing this mutant promoter ( Figure 4 ) . Residual decompaction was observed at the 3-fold stage , which may reflect transcriptional activity . This result suggests that PHA-4 is a critical contributor to large-scale decompaction of myo-2 , especially at early developmental stages . Conversely , we examined pseudo-chromosome arrays bearing mutant pax-1 mutA promoters , which were no longer transcribed but which bound PHA-4 . These arrays became decompacted despite the absence of productive transcription ( inside versus outside the pharynx ( p<0 . 0001 ) ) , whereas pax-1 mutP arrays bearing a mutated PHA-4 binding site did not ( Figure 4B ) ( smaller inside vs . outside the pharynx p = 2×10−14 ) . This result suggests that productive transcription is not essential for decompaction , and that PHA-4 association is sufficient . To test this idea more stringently , we created arrays bearing three repeats ( 3X ) of a PHA-4 binding site derived from pax-1 , but lacking additional promoter sequences . The 3X repeats were sufficient for PHA-4::YFP recruitment to the pseudo-chromosome and caused large-scale decompaction ( Figure 4C ) . These data reveal that PHA-4 binding , more than ongoing transcription , induces large-scale reorganization of chromatin in developing C . elegans embryos . PHA-4 is expressed broadly , including the pharynx , intestine , rectum , somatic gonad and some neurons [9] , [10] , [14] , [16] , [70] , yet PHA-4 targets are activated in discrete cell-types . For example , pax-1 is expressed in marginal cells but not in the intestine ( Figure 1 ) . We wondered if the discriminate activation of downstream targets could be explained by regulated binding of PHA-4 . PHA-4 binding was surveyed in a transgenic line carrying the pax-1 mutA promoter at three developmental stages ( bean , comma and 1 . 5-fold ) in one mid-section focal plane . Pharyngeal binding was detected in ∼67% of embryos at the bean stage ( 10/15 ) , ∼58% at the comma stage ( 7/12 ) , and ∼76% at the 1 . 5-fold stage ( 13/17; average 68% ) . By contrast , binding was almost never detected in the intestine at any stage ( <1%; 0/44 embryos counted; additional embryos surveyed but not counted; Figure 5A ) . Similar results were observed with arrays bearing myo-2 ( data not shown ) . An optical section through a 1 . 5-fold embryo sampled approximately 10 pharyngeal nuclei and 10 intestinal nuclei , indicating that the differential association of PHA-4 did not reflect different numbers of nuclei in each organ . Does regulated binding lead to differential PHA-4 activity in disparate tissues ? To answer this question we induced ectopic PHA-4 using a heat-shock promoter in transgenic lines bearing pax-1::GFP . HS::PHA-4 induced widespread expression of pax-1::GFP in many cells . However , we did not observe pax-1::GFP in the developing intestine ( 0/50 ) ( Figure 5B ) . This absence did not reflect variable PHA-4 expression , since antibody staining demonstrated that PHA-4 was expressed in intestinal cells equivalently to other tissues after heat shock ( Figure 5C ) . We detected no ectopic expression of the GFP reporter in non-heat shocked embryos ( Figure 5B ) , nor did we observe ectopic expression when we tested HS::pha-4DeltaDBD [9] , which lacked the DNA binding domain ( data not shown ) . These findings indicate that PHA-4 binding to pseudo-chromosome arrays limits PHA-4 activity , and that both binding and activity are sensitive to the cellular environment . This conclusion agrees with previous observations that HS::PHA-4 can induce embryonic cells to convert to a pharyngeal fate , but that the intestine is immune to ectopic PHA-4 [9] . To begin to understand the selective binding of PHA-4 in different cell types , we conducted a small RNAi screen for nuclear factors that modulate PHA-4 binding to target promoters . We used SM1634 carrying a mutant pax-1 promoter because pax-1-containing arrays typically bound PHA-4::YFP in fewer pharyngeal cells than myo-2-containing arrays ( data not shown ) . We surveyed genes involved in chromatin modification such as histone demethylation , methylation , acetylation and RNA interference . Given the proximity of the pseudo-chromosomes to the nuclear lamina , we also tested genes involved in nuclear envelope structure and function . We counted the number of nuclei with bound PHA-4::YFP in a section that passed through the pharynx and intestine of comma to 1 . 5 fold embryos . Of 28 genes surveyed , emr-1/Emerin had the most dramatic effect on PHA-4 binding ( Figure 6 ) . In the control , almost half of embryos had at least one nucleus with PHA-4::YFP bound to the pax-1 promoter , with an average of 17% pharyngeal nuclei bound within an embryo ( Figure 6B and 6C ) . Inactivation of emerin lead to a large increase in the number of pharyngeal nuclei with bound PHA-4::YFP , to ∼60% ( Figure 6A and 6B ) . Although EMR-1 is widely expressed in all embryonic tissues [36] , we observed binding only in the pharynx and not in the intestine of emr-1 ( RNAi ) embryos ( 1 of 88 embryos ( 1 . 25% ) in three experiments ) . These results reveal that the nuclear lamina interferes with binding of PHA-4::YFP to its targets within pharyngeal cells , but that additional processes function in the intestine . emr-1 ( RNAi ) strongly lowered expression of EMR-1 protein ( Figure S6 ) , and promoted pseudo-chromosome decompaction compared to wild-type embryos , raising the possibility that increased binding of PHA-4 in emr-1 ( RNAi ) embryos could be a consequence of increased accessibility . 32% ( 28/88 ) of emr-1 ( RNAi ) embryos had decondensed arrays compared to 17% ( 9/52 ) for wild-type ( Figure 6A and 6D ) . To explore the role of decompaction , we examined other genes for effects on pseudo-chromosome morphology and PHA-4::YFP binding . RNAi against 6 additional genes caused a global de-condensation of pseudo-chromosomal arrays at the comma to 1 . 5-fold stages of embryogenesis ( lem-3 , zyg-12 , lin-59 , set-1 , met-2 , ergo-1 Figure 6A and 6D ) . The arrays in these embryos appeared more distended and were brighter than wild-type embryos suggesting increased CFP::LacI expression . The de-condensation of the pseudo-chromosomes was not restricted to a specific tissue , but was observed in most nuclei in an optical section across the embryo . RNAi against set-1 , a gene encoding a potential SET-domain methyltransferase [71] , caused global decompaction in 19 of 45 embryos ( 40% , Figure 6D ) . The decompaction observed in set-1 ( RNAi ) embryos was not surprising given that set-1 has been implicated in transgene silencing [72] . However , set-1 ( RNAi ) embryos did not lead to increased PHA-4::YFP association ( Figure 6B and 6C ) . The most dramatic effect was observed for met-2 , a histone H3 lysine 9 dimethyltransferase that is homologous to human SETDB1 [73][74] . Arrays appeared more decompacted in met-2 ( RNAi ) nuclei , and a greater proportion of arrays were decondensed compared to those in wild-type embryos ( 77% , 28/36 , Figure 6A , 6D , and 6E ) . Decompaction by reduced met-2 had some effect on PHA-4 binding since a greater proportion of met-2 ( RNAi ) embryos had PHA-4::YFP localized to pseudo-chromosomes ( Figure 6C ) . However , within those embryos , only 29% of pharyngeal nuclei had PHA-4::YFP bound to pseudo-chromosomes , and this difference was not statistically significant from control . ( Figure 6A and 6C ) . These data indicate that general decompaction may influence PHA-4::YFP association , but that Emerin likely modulates PHA-4::YFP binding by additional mechanisms as well .
The transparency of C . elegans enables analysis of selector gene behavior in living embryos . Our characterization of PHA-4 and its target genes revealed regulated association with target promoters , which induced extensive chromatin decompaction in selected cells . We note that these events were visualized by single-cell analysis and would not have been detected by biochemical approaches such as chromatin immunoprecipitation . Importantly , the expression of pharyngeal reporter constructs embedded within complex DNA sequences , with few exceptions , mimics expression of the endogenous cognates , as detected by in situ stains [66] , [76] . Thus , the bulk of gene regulatory processes are preserved in the arrays . Large-scale decompaction of chromatin by a selector gene , which , to our knowledge , has not been observed previously , is consistent with observations regarding pha-4/Fox orthologues in other organisms . In breast cancer cells , global location analysis previously revealed that FoxA1 bound many regions located >50 Kb from a transcription start site [77] . FoxA1 induced both local effects , such as chromatin remodeling and transcription factor recruitment [54] , but also long-range effects , such as physical interactions between enhancers and promoters [77] . In S . cerevisiae , Fkh1/Fox controls donor preference during mating-type switching [55] . Fkh1 promotes recombination for loci separated by 50 Kb and does so without altering transcription or local chromatin [78] . These observations suggest Fox factors in diverse organisms contribute to long-range interactions between distant loci . In our system , we estimate roughly one PHA-4 target promoter per 25 Kb of DNA within the pseudo-chromosomes . This number derives from ∼200 copies of target promoter ( qPCR , data not shown ) embedded in arrays of ∼5–7 Mb [79] . At endogenous loci , PHA-4::GFP associates with 4350 sites in the embryonic genome , a surprisingly high density of binding sites [19] . The distance observed in C . elegans is comparable to those in the yeast and mammalian studies , suggesting that large-scale , Fox-mediated chromatin re-organization might operate in all three organisms . The association of PHA-4::YFP with pseudo-chromosomes was not constitutive , but responded to the cellular milieu in two ways . First , within the pharynx , PHA-4::YFP binding was restricted by EMR-1/emerin . EMR-1/emerin resides at the nuclear lamina , which suggests that tethering of pharyngeal genes or trans-acting factors at the nuclear periphery may modulate PHA-4 binding . We note that the nuclear lamina appears normal after emr-1 ( RNAi ) , and affected embryos are healthy and viable ( This study and [43] ) . We speculate that the loss of emerin may have subtle effects on tethering or formation of heterochromatin , signaling pathways or lamina-associated proteins that alter gene activity [80] . Second , in the intestine , PHA-4 binding to pax-1 and myo-2 was inhibited completely , and inhibition was not relieved by emr-1 ( RNAi ) . HS::pha-4 cannot activate pax-1::GFP within intestinal cells ( this study ) or convert nascent intestinal cells to a pharyngeal fate [9] , indicating PHA-4 functions poorly in this embryonic tissue . By contrast , ubiquitous expression of the C . elegans MyoD homolog hlh-1 induces the body-wall muscle program throughout the embryo , including developing intestinal cells [48] , [81] . We suggest that the limited activity of PHA-4 within the intestine may reflect the inability of PHA-4 to associate with its pharyngeal target genes . This lack of association in the intestine may reflect the presence of gut-specific repressive systems that block pharyngeal gene activation in the intestine , or the absence of appropriate cofactors and coactivators . What is the nature of PHA-4-induced chromatin restructuring ? The global decompaction we observe is consistent with a disordered structure , such as decondensation by loss of nucleosomes and/or reconfiguring of chromatin into loops or coils [82] . Although nucleosome loss can be associated with transcription [83] , our data suggest that the effect of PHA-4 is independent of productive transcription . Arrays bearing the mutA promoter or 3X PHA-4 binding site repeats recruit PHA-4::YFP and undergo decompaction , in the absence of GFP production . In Drosophila , nucleosomes are lost rapidly at heat-shock loci prior to transcription , and this loss extends across several kilobases upstream and downstream of the activated gene [84] . Transcription-independent decondensation of chromatin might be required to “clear the way” for RNA Pol II , enabling cells to activate gene expression rapidly and respond promptly to developmental and environmental cues . An interesting feature of pax-1 expressing cells is that they share a lineage relationship . We identified 11 of the 14 pax-1::GFP+ cells unambiguously , and found that each of these cells derived from the posterior daughter of the penultimate cell division ( “px” cells; Figure S1 ) . For example , ABaraaapapa generates a marginal cell that expresses pax-1 . Previous studies have shown that C . elegans embryos are patterned according to antero-posterior ( A-P ) cell divisions in which pairs of A-P siblings are distinguished by high ( anterior ) or low ( posterior ) levels of nuclear POP-1 , a TCF transcription factor ( [85] , [86] reviewed in [66] ) . Loss of POP-1 asymmetry alters cell fate decisions , suggesting transcriptional regulation by POP-1 confers anterior or posterior identity after each cell division [85] , [86] . However , few transcriptional targets of POP-1 are known . We considered an appealing model that POP-1 might regulate pax-1 transcription directly during the penultimate cell division and thereby contribute to A–P fate distinctions . However , none of the cis-regulatory sites we identified are a good fit with the canonical TCF binding site G ( A/T ) ( A/T ) CAAAG [87] . Thus , the relationship between pax-1 and A-P specification remains a mystery . Our promoter analysis identified four regulatory elements that establish pax-1 expression in fourteen pharyngeal cells . The first was an enhancer element likely recognized by PHA-4 and defined by D6 . PHA-4 can bind this sequence in vitro [2] and in vivo ( this study ) . Moreover , this site is required for pharyngeal expression ( this study ) , and multimers of this sequence respond to PHA-4 in vivo [18] . This result supports the notion that many genes expressed within the pharynx are direct targets of PHA-4 [2] . Surprisingly , while mutation of the predicted PHA-4 binding site eliminated pax-1 expression within the pharynx , it also led to ectopic expression in non-pharyngeal cells such as epidermis . M05B5 . 2 and T05E11 . 3 are two additional PHA-4 target genes [2] , and these also exhibited epidermal expression when the PHA-4 site was mutated to random sequences ( J . Gaudet , pers . comm . ) . A likely possibility is that this site functions as a repression element in non-pharyngeal epithelia . PHA-4 is not expressed in the epidermis , leaving open the identity of the factor that represses epidermal expression . RNAi of the other C . elegans Fox genes did not result in ectopic expression in lines carrying the wildtype M05B5 . 2 reporter ( J . Gaudet , unpublished ) . This result suggests that multiple Fox proteins function redundantly to repress epidermal expression , or alternatively , that an unrelated protein acts through the predicted PHA-4 binding site . A second enhancer element defined by Delta16 contributes to pax-1 activation . The Delta16 region contains a match to a GATA-2 , 3 binding site ( AGATTA; [88] , [89] . However , mutation of AGATTA to CTGCAG does not inactivate pax-1 expression , suggesting this site is not recognized by a GATA factor ( J . S . , data not shown ) . We note that the sequence TTGAGA lost in Delta16 is half of a direct repeat , with a second copy located within Delta14 ( Figure 1 ) . Abutting Delta16 sequences , mutations Delta14 and Delta18 each lead to pax-1 expression in extra pharyngeal cells . These sequences carry an inverted repeat AGAGCT that is lost in Delta14 or Delta18 ( Figure 1 ) . Two additional elements , defined by Delta20 and Delta22/Delta24 , functioned negatively to restrict pax-1 expression . A direct repeat ( ACGGACCA ) lies within these sequences , with one copy entirely within Delta20 and a second spanning Delta22 and Delta24 . An appealing model is that PHA-4 promotes expression within the pharynx in combination with Delta16 sequences . The broad activation is refined by the repression elements embedded in Delta14/Delta18 and Delta20/Delta22/Delta24 . The combination of four cis-regulatory sites explains why pan-pharyngeal PHA-4::YFP can bind its target promoters , yet those targets become transcriptionally active in only a subset of pharyngeal cells and after PHA-4 is first expressed . We have demonstrated that the master regulator PHA-4 binds to its pharyngeal targets hours before the onset of gene expression . PHA-4 binding and activity is restricted in the intestine and negatively regulated by EMR-1/emerin in the pharynx . The association of PHA-4 with target promoters led to large-scale chromatin decompaction , which may facilitate chromatin-associated processes such as transcription . These in vivo results expand our understanding of PHA-4/FoxA function in driving pharyngeal transcriptional programs . Moving beyond the Nuclear Spot Assay , it will be interesting to investigate the binding and down-stream consequences of PHA-4 in its native environment , at endogenous loci .
Strains were maintained as described in [90] , at 20°C , and were provided by Caenorhabditis Genetics Center , which is funded by the NIH National Center for Research Resources ( NCRR ) , unless stated otherwise . Bristol N2 was used as the wild-type strain . The following mutation was used LGIV: cha-1 ( p1182 ) . For pax-1::GFP analysis the following transgenic strains were used: SM202 pxls2 ( pax-1::GFP + pRF4 ) , SM699 N2 ( pax-1::GFP + pRF4 ) , SM707 N2 ( pax-1 mutP-pro::GFP + pRF4 ) , SM658 N2 ( pax-1 mutA-pro::GFP + pRF4 ) , SM660 N2 ( pax-1 Delta14-pro::GFP ) , SM700 N2 ( pax-1Delta18-pro::GFP ) . For Heat Shock: SM259 pxEx ( HS::PHA-4 + pax-1::GFP + UL8::lacZ + pRF4 + 1 KB ladder + Herring Sperm DNA ) [9] . For the Nuclear Spot Assay ( NSA ) , the following strains were used: SM1560 cha-1 ( p1182 ) ; pxEx ( cha-1 + his-24pro::CFP::LacI + pha-4::yfp+ lacO + Herring Sperm DNA ) , SM1476 cha-1 ( p1182 ) ; pxEx ( cha-1 + htz-1pro::CFP::LacI + PHA-4::YFP +myo-2proWT + lacO + Herring Sperm DNA ) , SM1429 cha-1 ( p1182 ) ; pxEx ( cha-1 + htz-1pro::CFP::LacI + PHA-4::YFP +myo-2proWT + lacO + Herring Sperm DNA ) , SM1443 cha-1 ( p1182 ) ; pxEx ( cha-1 + htz-1pro::CFP::LacI + PHA-4::YFP +myo-2 mutP + lacO + Herring Sperm DNA ) , SM1444 cha-1 ( p1182 ) ; pxEx ( cha-1 + htz-1pro::CFP::LacI + PHA-4::YFP +myo-2 mutP + lacO + Herring Sperm DNA ) , SM1432 cha-1 ( p1182 ) ; pxEx ( cha-1 + htz-1pro::CFP::LacI + PHA-4::YFP + pax-1proWT + lacO + Herring Sperm DNA ) , SM1434 cha-1 ( p1182 ) ; pxEx ( cha-1 + htz-1pro::CFP::LacI + PHA-4::YFP + pax-1proWT + lacO + Herring Sperm DNA ) , SM1463 cha-1 ( p1182 ) ; pxEx ( cha-1 + htz-1pro::CFP::LacI + PHA-4::YFP + pax-1 mutP + lacO + Herring Sperm DNA ) , SM1628 cha-1 ( p1182 ) ; pxEx ( cha-1 + his-24pro::CFP::LacI + PHA-4::citrineYFP + pax-1Delta6proMut + lacO + Herring Sperm DNA ) , SM1564 cha-1 ( p1182 ) ; pxEx ( cha-1 + his-24pro::CFP::LacI + PHA-4::citrineYFP + pax-1 mutA + lacO + Herring Sperm DNA ) , SM1634 cha-1 ( p1182 ) ; pxEx ( cha-1 + his-24pro::CFP::LacI + PHA-4::citrineYFP + pax-1 mutA + lacO + Herring Sperm DNA ) , SM1523 cha-1 ( p1182 ) ; pxEx ( cha-1 + his-24pro::CFP::LacI + PHA-4::citrineYFP + 3X low affinity pha-4 site + lacO + Herring Sperm DNA ) . SM1876 cha-1 ( p1182 ) ; stIs10389 ( pha-4::gfp::3xFLAG ) ; pxEx ( cha-1 + htz-1pro::mCherry::LacI + M05B5 . 2 + lacO + Salmon testes DNA ) . Transgenic worms used for the Nuclear Spot Assay ( NSA ) were grown at 24°C on an E . coli OP50 lawn or on RNAi plates ( see below ) . For the pax-1::GFP cytoplasmic translational fusion construct ( BSEM74 ) : a 4 . 6 kb genomic SacI DNA fragment was cloned from K07C11 . 1 into pBluescriptIISK+ . The resulting plasmid was digested with NsiI , which is located within the pax-1 locus , and XbaI from the polylinker , to generate a 3 kb pax-1 fragment that was inserted into PstI/XbaI-digested pPD95 . 77 ( A gift from Dr . Andrew Fire ) . The resulting pax-1::GFP reporter contained approximately 2 . 4 KB upstream sequences , with GFP fused to pax-1 within the second predicted exon of pax-1 . The transcriptional pax-1 nuclear construct ( BSEM274 ) was made starting with the cytoplasmically-expressed pax-1::GFP translational construct , we created a transcriptional fusion by removing all coding sequence . We performed inverse PCR using primers containing BglII tails that flanked the region to be deleted . The linear PCR product was then digested with BglII and re-ligated , placing the GFP translational start site at the same position as the one removed for pax-1 . PCR products for injection were generated using pax-1 5′ as the forward primer and pax-1 3′ as the reverse primer . The transcriptional fusion construct was modified for use in the scanning mutagenesis . A 1 . 2 kB Bst1071I fragment was removed from BSEM274 , and replaced with a 1 . 9 kb Bst1107I/ApaI fragment from pAP . 10 that extends from within the GFP coding region through the unc-54 poly A addition site was removed . This generates a pax-1::GFP transcriptional fusion with the coding sequence for histone H2B fused to the 3′ end of GFP . 276 , 277 , 279 , 280 . pha-4::citrineYFP for the nuclear spot assay was created using QuickChange site directed mutagenesis ( Stratagene , #200519 ) . Two mutations , V68L and Q69M ( 5′-GTT-CAA-3′ mutated to 5′-CTT-ATG-3′ ) , were introduced into the YFP sequence of the pha-4::YFP ( SEM962 ) [11] to convert YFP into citrineYFP [91] . The following primers were used: YFP FW Citrine 5′-GTCAC TACTTTCGGTTATGGTCTTATGTGCTTCGCCAGATACCCAGATC-3′ and YFP RV Citrine 5′GATCTGGGTATCTGGCGAAGCACATAAGACCATAACCGAAAGT AGTGAC-3′ . 3X low affinity PHA-4 binding site oligos were designed as in [18] but without restriction sites flanking the 3 tandem binding sites and were created to have an overhang to facilitate repeat formation in the array . The oligos are: Top 5′-CTACTATTTGTCCCTACTATTTGTCCCTACTATTTGTCC-3′ Bottom 5′ GGGACAAATAGTAGGGACAAATAGTAGGGACAAATAGTA-3′ ( underlined are the PHA-4 binding sites ) . Oligos were diluted to a concentration of 2 mg/ml and heated to 95°C for 3 minutes . The temperature was dropped 0 . 10/sec until 20°C was reached to hybridize the oligos . Reporter constructs were injected into the germ line of hermaphrodites and stable F2 transgenic roller lines examined . All reporter constructs were injected at 0 . 5 ng/ml as PCR fragments . Low concentrations of PCR products circumvented artificial expression in pharyngeal cells that has been observed with plasmid constructs [92] . We used pRF4 ( rol-6 ( su1006 ) ) as a co-injection marker [93] . The injection mix also included complex DNA ( salmon sperm DNA , 1 kb ladder ) up to 100 ng/ml , to prevent silencing [59] . Antibody staining was performed as described previously [11] . The primary antibodies used were anti-GFP rabbit IgG fraction at 1∶1000 ( Molecular Probes ) , anti-PHA-4 PAb at 1∶1000 [94] , the anti-intermediate filament ( a-IF ) at 1∶3 that recognizes pharyngeal marginal cells [95] , and the monoclonal antibody J126 , from Dr . Susan Strome , was used at 1∶30 to detect intestinal cells . All constructs for the scanning mutagenesis were constructed using an inverse PCR strategy . For each mutant , we used a specific pair of primers that flank the 10 bp region to be altered . These primers each carry 5′ tails that contain a restriction site ( PstI or ClaI ) . Following inverse PCR ( BSEM279 as template ) , the linear PCR product was digested with the appropriate restriction enzyme ( PstI or ClaI ) and re-ligated . Each resulting reporter plasmid contains the restriction site plus a variable number of base pairs in place of the 10 bp of wild-type sequence . For injection , PCR products were generated from each mutant plasmid . All constructs were sequenced to confirm the predicted sequence . Transgenic lines for the Nuclear Spot Assay were as follows: no target control ( SM1560 ) , 3X low affinity pha-4 binding sites ( SM1523 ) , myo-2 wild-type promoter ( SM1476 , SM1429 ) bearing two high affinity PHA-4 binding sites [2] , myo-2 promoter bearing mutagenized FoxA sites ( SM1443 , SM1444 ) [2] , pax-1 wild-type promoter ( SM1432 , SM1434 ) , pax-1 promoter with a mutagenized FoxA site ( SM1463 , SM1628 ) and pax-1 promoter with a mutagenized positive regulator site ( SM1564 , SM1634 ) ( see below ) . SM1560 was created by injecting cha-1 ( p1182 ) worms with Xho1-linearized pha-4::citrineyfp plasmid ( bSEM1045 ) ( 1 ng/µl ) , his24promoter::CFP::LacI PCR product [63] ( 2 . 5 ng/µl ) , a 10 kb Sph1/Kpn1 fragment from lacO multimeric plasmid pSV2-dhfr-8 . 23 ( 3 ng/µl ) [96] , cha-1 plasmid ( RM527P , a gift from J . Rand ) linearized with Apa1 ( 2 ng/µl ) for rescue , and sheared herring sperm DNA to make 100 ng/µl total DNA . For SM1476 and SM1429 , 499 bp of the endogenous myo-2 promoter upstream of the start codon was used in addition to the components listed for SM1560 with one difference , CFP::lacI expression was driven by the htz-1 promoter ( BSEM995 ) [63] . SM1443 and SM1444 were created similar to SM1476 and SM1429 but with a myo-2 promoter bearing two mutated PHA-4 binding sites [2] . For SM1434 a 240 ( bp ) fragment of the pax-1 promoter upstream of the start codon was used ( the fragment contains one PHA-4 binding site ( TGTTTGC ) ) . SM1463 carried an altered version of the 240 ( bp ) pax-1 promoter in which the PHA-4 binding site was mutated from TGTTTGC to ATCGATT ( MutP ) . Both SM1463 and SM1434 were injected with htz-1pro::CFP::LacI . For SM1564 and SM1634 a positive regulator site −40 to −50 upstream of the TSS was mutated from TTGAGATTAA to CAATCGATTG . SM1876 was created by injecting SM1754 cha-1 ( p1182 ) ; stIs10389 ( pha-4::gfp::3xFLAG ) worms with cha-1 , a 440 ( bp ) M05B5 . 2 promoter fragment [2] , and htz-1pro::mCherry::LacI that has a premature stop codon at the end of the mCherry sequence , thus failing to make any mCherry::LacI . SM1876 was used to examine whether decompaction reflected an artificial interaction between LacI and PHA-4 . Nuclear spot assays were performed as described previously [11] , [50] , [51] , [63] , with the following modifications: sequential scan images were acquired using the Andor Revolution XD microscopy system ( ≤16E stages ) . For later stages , images were acquired using an Olympus FluoView FV1000 confocal microscope ( for pax-1 mutA ) or a Leica DM RXE confocal ( for everything else ) . To determine copy number , worms were grown at restrictive temperature for cha-1 ( p1182 ) ( 25° ) , and treated with bleach to synchronize embryos . Four 10 cm OP50 plates of moving , nonCha-1 L3 animals were harvested for DNA isolation by phenol chloroform extraction and ethanol precipitation . qPCR was performed for promoter regions and normalized to act-1 using a LightCycler PCR machine with LightCycler FastStart DNA MasterPlus SYBR Green 1 kit ( Roche ) for quantitation . qPCR indicated a copy number of ≤200 for each promoter . Gravid mothers were dissected and embryos collected in a PCR tube . Heat-shock was administered in a PCR machine . Embryos were initially incubated at 20°C for 75 min . After the initial incubation , the temperature was raised gradually to 33°C at a rate of 0 . 1°C/second . Embryos were then incubated at 33°C for 30 minutes . Following heat shock , the temperature was gradually lowered to 20°C at a rate of 0 . 1°C/second , and embryos incubated at 20°C for 5 hours . Perkin Elmer Volocity was used to calibrate images for true X and Y pixel dimensions to ensure accurate spatial measurements . Classifiers were designed to select CFP::LacI areas and PHA-4::YFP+ cells using an intensity threshold . The CFP::LacI classifier included separation of touching objects , removal of noise and exclusion of objects smaller than 0 . 25 micron2 . The PHA-4::YFP classifier was modified to remove noise . CFP::LacI areas within PHA-4::YFP+ cells were considered “inside the pharynx” and the remainder as “outside the pharynx . ” Proofreading of selections was performed blind by comparing measurements with images . Area measurements ( INT Area ( micron2 ) *10 ) were analyzed using Cox regression models to evaluate differences in chromosome area with location ( inside the pharynx versus outside ) , developmental stage , or transgenic line . While Cox regression models were originally developed for analyzing survival data , their semi-parametric nature made them suitable for analyzing data following a non-standard distribution that was difficult to capture parametrically . RNAi by bacterial feeding was performed similarly to [15] . HT115 bacteria [97] expressing dsRNA for gfp , spr-5 , rbr-2 , npp-11 , emr-1 , nhr-60 , ima-3 , lem-3 , zyg-12 , lmn-1 , lin-49 , lin-59 , set-17 , set-16 , set-2 , set-1 , met-2 , hda-3 , hda-4 , tsn-1 , prg-1 , sago-2 , prg-2 , csr-1 , top-1 , ergo-1 , chd-3 , tam-1 , lin-35 , or hil-7 were grown in liquid cultures for 8 hours and seeded onto plates containing 8 mM IPTG ( Sigma ) and 50 g/ml Carbenicillin ( Sigma ) . All RNAi clones were derived from the Ahringer library [98] . The emr-1 clone was validated by sequencing using pPD129_for 5′-GAGTGAGCTGATACCGCTCG-3′ and pPD129_rev 5′-CACGACGGTGTATTTCGACGGC-3′ primers at the Dana-Farber/Harvard Cancer Center DNA Resource Core . Adult SM1634 worms were bleached and ∼50 embryos were placed on RNAi plates ( Po ) . For every experiment , F1 progeny embryos from 5 6 cm plates were collected by bleaching and analyzed , and each experiment was repeated at least twice . Images were acquired from live embryos using an Olympus FluoView FV1000 confocal microscope and DeltaVision RT Deconvolution System and SoftWoRx software ( Applied Precision ) . A multitrack setting was used to acquire separate CFP and YFP images from slices through the pharynges of comma to 1 . 5fold-stage embryos . Embryos were scored for general non-tissue specific extrachromosomal array decompaction and for the number of nuclei containing PHA-4-::YFP colocalized with CFP::LacI .
|
Central regulators of cell fate establish the identity of cells by direct regulation of large cohorts of genes . In Caenorhabditis elegans , foregut ( or pharynx ) identity relies on the FoxA transcription factor PHA-4 , which activates different target genes in different cellular environments . An outstanding question is how PHA-4 distinguishes between target genes for appropriate transcriptional control . Here we examine PHA-4 interactions with target promoters in living embryos and with single-cell resolution . While PHA-4 was found throughout the digestive tract , binding and activation of pharyngeally expressed promoters was restricted to a subset of pharyngeal cells and excluded from the intestine . An RNAi screen identified emerin ( emr-1 ) as a negative regulator of PHA-4 binding within the pharynx . Upon promoter association , PHA-4 induced large-scale chromatin de-compaction , which , we hypothesize , facilitates promoter access . Our results reveal two tiers of PHA-4 regulation . PHA-4 binding is prohibited in intestinal cells and is limited in the pharynx by the nuclear lamina component EMR-1/emerin . The data suggest that association of PHA-4 with its targets is a regulated step that contributes to promoter selectivity during organ formation . We speculate that global re-organization of chromatin architecture upon PHA-4 binding promotes competence of pharyngeal gene transcription and , by extension , foregut development .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology/embryology",
"developmental",
"biology",
"molecular",
"biology/transcription",
"initiation",
"and",
"activation",
"developmental",
"biology/cell",
"differentiation",
"developmental",
"biology/organogenesis"
] |
2010
|
Dynamic Chromatin Organization during Foregut Development Mediated by the Organ Selector Gene PHA-4/FoxA
|
Members of the APOBEC3 family of deoxycytidine deaminases counteract a broad range of retroviruses in vitro through an indirect mechanism that requires virion incorporation and inhibition of reverse transcription and/or hypermutation of minus strand transcripts in the next target cell . The selective advantage to the host of this indirect restriction mechanism remains unclear , but valuable insights may be gained by studying APOBEC3 function in vivo . Apobec3 was previously shown to encode Rfv3 , a classical resistance gene that controls the recovery of mice from pathogenic Friend retrovirus ( FV ) infection by promoting a more potent neutralizing antibody ( NAb ) response . The underlying mechanism does not involve a direct effect of Apobec3 on B cell function . Here we show that while Apobec3 decreased titers of infectious virus during acute FV infection , plasma viral RNA loads were maintained , indicating substantial release of noninfectious particles in vivo . The lack of plasma virion infectivity was associated with a significant post-entry block during early reverse transcription rather than G-to-A hypermutation . The Apobec3-dependent NAb response correlated with IgG binding titers against native , but not detergent-lysed virions . These findings indicate that innate Apobec3 restriction promotes NAb responses by maintaining high concentrations of virions with native B cell epitopes , but in the context of low virion infectivity . Finally , Apobec3 restriction was found to be saturable in vivo , since increasing FV inoculum doses resulted in decreased Apobec3 inhibition . By analogy , maximizing the release of noninfectious particles by modulating APOBEC3 expression may improve humoral immunity against pathogenic human retroviral infections .
Millions of years of co-evolution of retroviruses and mammalian hosts has led to the emergence of retroviral host restriction factors , some of which were discovered following major efforts to understand key steps in the HIV life cycle . Most of these genes , such as TRIM5α [1] , Tetherin/Bst2 [2] and SAMHD1 [3] , restrict retroviruses in the infected cell . In contrast , members of the APOBEC3 family of deoxycytidine deaminases are distinguished by their ability to inhibit retroviruses in the next target cell . In cell culture , co-transfection of APOBEC3 expression plasmids with retrovirus molecular clones does not decrease virus output , but the infectivity of the resulting virions is dramatically decreased [4]–[5] . These APOBEC3-containing virions are fusion-competent , but encounter post-entry blocks from early reverse transcription to integration [6]–[7] , with G-to-A hypermutation of nascent reverse transcripts observed in most , but not all [8]–[11] retrovirus infections . Notably , while in vivo studies have largely focused on G-to-A hypermutation as a read-out of APOBEC3 function [12]–[14] , the biological relevance of APOBEC3-mediated reduction of virion infectivity remains unclear . In fact , it is currently unknown whether high viral output with reduced infectivity can be detected in vivo , since multiple rounds of replication with Apobec3 restriction will ultimately result in decreased total virus titers [15]–[16] . Investigating the impact of APOBEC3 restriction in vivo is logistically difficult in humans due to potential redundancy in antiretroviral activities of seven human APOBEC3 members ( APOBEC3A , B , C , D , F , G and H ) ( reviewed in [17] ) . Moreover , APOBEC3 activity is likely most relevant immediately following viral transmission , but such biological samples are very difficult to obtain from pathogenic human retrovirus infections . In contrast , mice encode a single Apobec3 gene ( mA3 ) [4]–[5] that could be genetically disrupted . While mA3-deficient mice are physiologically normal [18] , they proved more susceptible to pathogenic murine retroviral infections that include Mouse Mammary Tumor virus [8] , Moloney Murine Leukemia Virus [19] and Friend retrovirus ( FV ) complex [20]–[21] . These studies provide a springboard for investigating the immunological impact of mA3 restriction in vivo . FV causes severe splenomegaly and erythroleukemia in adult immunocompetent mice , and resistance and susceptibility to FV have been mapped to a variety of genes [22]–[23] . One of these genes , Recovery from Friend retrovirus gene 3 ( Rfv3 ) , influences the recovery of mice from viremia by promoting a more potent neutralizing antibody ( NAb ) response [24]–[25] . C57BL/6 ( B6 ) mice recover from viremia , develop stronger NAb responses and are Rfv3 resistant , while BALB/c , A . BY and A/WySn strains have persistent viremia , develop weaker NAb responses and are Rfv3 susceptible . Our group and others demonstrated that the B6 mA3 gene acts as the classical Rfv3 resistance gene , promoting stronger NAb responses and facilitating recovery from FV viremia , infection , and disease in ( B6×BALB/c ) F1 , ( B6×A . BY ) F1 and ( B6×A/WySn ) F1 mice [20] , [26]–[27] . In addition , the B6 mA3 gene restricted acute FV replication in immune compartments [20]–[21] , [27]–[28] . Acute FV inhibition was associated with significantly higher mA3 mRNA expression and splicing differences in Rfv3 resistant ( B6 ) compared to Rfv3 susceptible strains ( BALB/c , A . BY , A/WySn ) [10] , [21] , [26] , [29] . However , the mechanism through which B6 mA3 promotes FV-specific NAb responses remains unknown . The APOBEC3 genes are evolutionarily related to Activation-Induced Deaminase , a B-cell specific enzyme that is critical for antibody affinity maturation and class-switching [30] . Thus , the identification of mA3 as Rfv3 led to the immediate hypothesis that mA3 may directly influence antibody development [20] . However , hapten immunization studies revealed that B6 mA3 influenced antibody affinity maturation only in the context of FV infection [28] . Thus , the underlying mechanism for the mA3/Rfv3 phenotype does not involve a direct effect of mA3 on B cell function . In fact , decreased immune dysfunction was found to be a critical component of how B6 mA3 promotes NAb responses [27]–[28] . We therefore hypothesized that mA3 influences NAb responses by promoting the release of noninfectious virions in vivo [28] , driving NAb responses without eliciting pathology . The mechanism for the mA3/Rfv3 phenotype may have implications for improving humoral immunity against human retroviruses , particularly against HIV-1 . However , HIV-1 encodes Vif , which promotes the degradation of the human homologues APOBEC3G ( hA3G ) and APOBEC3F ( hA3F ) [31]–[34] . Surprisingly , despite the action of Vif , hA3G/hA3F-mediated G-to-A hypermutation was detected in HIV-1 sequences from clinical specimens [12]–[14] . These findings could in part be due to the emergence of defective Vif alleles [35] . However , high hA3G mRNA levels in primary cells or tissues were also associated with lower HIV-1 viral loads [36]–[37] . In rhesus macaques infected with SIV , rhesus macaque A3G ( rhA3G ) levels in colonic biopsies following mucosal vaccination also correlated with set-point plasma viral loads [38] . Thus , hA3G/hA3F may be induced to levels that saturate endogenous levels of Vif . This is consistent with in vitro observations that show inhibition of wild-type HIV-1 with increasing hA3G transfection levels [4]–[5] . Unfortunately , obtaining direct evidence that innate hA3G/hA3F restriction is saturable in humans is not feasible . We therefore evaluated the in vivo antiviral activity and saturability of B6 mA3 in the context of FV infection of mice . The results provide long-sought insights into a fundamental APOBEC3 restriction phenotype that may have important implications for HIV-1 vaccine research .
( B6×BALB/c ) F1 mice ( genotype Rfv3r/s ) are highly susceptible to FV infection but recover from plasma viremia by 28 days post-infection ( dpi ) due to the dominant , B6-encoded Rfv3 resistance gene [20] , [26] ( Figure S1A in Text S1 ) . In contrast , the majority of ( B6 mA3−/−×BALB/c ) F1 mice ( genotype Rfv3−/s ) do not survive to 28 dpi , and those that survive display elevated plasma viremia , consistent with identity between Rfv3 and mA3 [20] , [26] ( Figure S1B in Text S1 ) . We previously reported that the significant survival disadvantage of ( B6 mA3−/−×BALB/c ) F1 mice was linked to higher plasma viremia during acute infection ( 7 dpi ) , quantified on susceptible Mus dunni cells using an FV envelope-specific monoclonal antibody to detect foci of infectivity [39] . Thus , the Mus dunni assay measures infectious viremia . In contrast , a quantitative RT-PCR assay measuring total viral RNA copies ( specifically , the F-MuLV helper virus component , as described in Materials and Methods ) does not distinguish between infectious and noninfectious virions [26] , [40] . To determine if B6 mA3 affected the relative amount of infectious virions released in vivo , we measured the ratio of virus titers obtained from both assays in plasma from ( B6×BALB/c ) F1 mice infected with 140 SFFU of FV ( Figure 1 ) . Interestingly , plasma samples from ( B6 mA3+/+×BALB/c ) F1 mice that had 10-fold lower infectious viremia titers at 7 dpi compared to ( B6 mA3−/−×BALB/c ) F1 mice ( Figure 1A ) had equivalent total viral RNA loads ( Figure 1B ) . Thus , the fraction of infectious viral particles was significantly higher in mice without a functional B6 mA3 gene ( Figure 1C ) . By setting the mean virion infectivity of ( B6 mA3−/−×BALB/c ) F1 mice to 100% , the relative infectivity of plasma virions from ( B6 mA3+/+×BALB/c ) F1 mice averaged only 22% ( Figure 1D ) . Thus , B6 mA3 restriction resulted in approximately 5-fold higher levels of noninfectious virions at 7 dpi . Higher proportions of noninfectious particles were also observed in ( B6 mA3+/+×A . BY ) F1 versus ( B6 mA3−/−×A . BY ) F1 mice ( Figure S2 in Text S1 ) . In the absence of HIV-1 Vif , hA3G and hA3F mediate high levels of G-to-A hypermutation in the minus strand of viral DNA , disrupting open reading frames and effectively inactivating HIV-1 . In contrast , mA3 did not appear to induce G-to-A hypermutation against FV , MMTV and MLV [8]–[11] , but these data were obtained either from cells infected in vitro or from bulk infected tissues . To investigate the mechanism of B6 mA3 restriction of plasma virions released during acute infection in vivo , we monitored FV sequence evolution in plasma from infected ( B6 mA3+/+×BALB/c ) F1 and ( B6 mA3−/−×BALB/c ) F1 mice ( Figure 2A ) . Partial env sequences amplified from the FV inoculum stock phylogenetically clustered with each other ( Fig . 2B , left panel ) , supporting their authenticity as reference sequences . Mutations that were already present in the inoculum quasispecies ( Figure 2B , right panel ) were then excluded from FV mutational analyses in infected mice . Multiple FV env sequences were obtained from 7 dpi plasma of ( B6 mA3+/+×BALB/c ) F1 and ( B6 mA3−/−×BALB/c ) F1 mice . Relative to the FV inoculum sequences , the cumulative mutational load ( Figure 2C ) , G-to-A substitutions ( Figure 2C ) , and the continuity of the envelope open reading frames ( Figure S3 in Text S1 ) in plasma viral RNA from ( B6 mA3+/+×BALB/c ) F1 and ( B6 mA3−/−×BALB/c ) F1 mice did not significantly differ from each other . Thus , the reduction in the infectivity of 7 dpi plasma virions from ( B6 mA3+/+×BALB/c ) F1 mice ( Figure 1C ) was likely not due to disproportionately mutated viral genomes . Newly formed FV reverse transcripts in target cells following infection with plasma virions were next evaluated using approaches to bias for detection of G-to-A mutations [11] , [41]–[42] . When compared to FV sequences from the inoculum and the corresponding 7 dpi plasma samples , similar mutation frequencies and total G-to-A substitutions were observed in ( B6 mA3+/+×BALB/c ) F1 and ( B6 mA3−/−×BALB/c ) F1 mice ( Figure 2D ) . However , when these G-to-A substitutions were partitioned into mA3-associated dinucleotide preferences , the ( B6 mA3+/+×BALB/c ) F1 strain was associated with the detection of GG→AG and GA→AA mutations ( Figure 2D ) . Thus , we could detect signatures of mA3-associated G-to-A mutations from reverse transcripts generated from acute B6 mA3+ plasma virions . However , even with techniques that significantly biased for the detection of such reverse transcripts , the G-to-A substitution frequency obtained for FV ( 0 . 14% ) was at least 10-fold lower than that observed for HIV-1 ΔVif , which ranges from 1 . 3 to 6 . 5% [14] . Together , these findings suggest that mA3-mediated deamination plays a very minor role in restricting FV infection in vivo . The lack of mA3-associated G-to-A substitutions in reverse transcripts from acute plasma virions of ( B6 mA3+/+×BALB/c ) F1 mice argued in favor of a deamination-independent mechanism of mA3 inhibition . Human A3G can non-enzymatically impair an early step in HIV-1 reverse transcription involving the generation of strong stop DNA [7] , [43] . Accordingly , we quantified the levels of newly formed early ( R-U5 or strong-stop DNA ) and late ( R-gag ) reverse transcripts following single-round infection of target cells with 7 dpi plasma FV virions ( Figure 2A ) . These studies revealed a 9-fold decrease in early reverse transcripts from ( B6 mA3+/+×BALB/c ) F1 plasma virions compared to F1 mice lacking B6 mA3 ( Figure 2E , left panel ) . No further decrease was observed in late reverse transcripts ( Figure 2E , right panel ) . Thus , the post-entry block conferred by B6 mA3 on FV plasma virions occurred primarily during the earliest stages of reverse transcription . We previously reported that B6 mA3 decreased FV infection levels in target cells that include erythroblasts and B cells in ( B6×A . BY ) F1 mice at 7 dpi [28] . FV infected cells were quantified by flow cytometry using a Glyco-Gag specific monoclonal antibody , MAb 34 [44] . We now extend this observation to ( B6×BALB/c ) F1 mice . As shown in Table 1 , the percentage of MAb 34+ cells was consistently lower in ( B6 mA3+/+×BALB/c ) F1 compared to ( B6 mA3−/−×BALB/c ) F1 strains in multiple cell subpopulations in the bone marrow and the spleen . Statistical significance was achieved with bone marrow erythroblasts and splenic B , T and dendritic cells . These findings revealed that at 7 dpi , B6 mA3 restriction of FV virion infectivity coincided with decreased FV infection of multiple target cells in vivo . To test whether B6 mA3 restriction was saturable in vivo , we infected ( B6 mA3+/+×BALB/c ) F1 and ( B6 mA3−/−×BALB/c ) F1 mice with titrated doses of FV , and measured infectious plasma viremia at 7 dpi ( Figure 3A ) . ( B6 mA3−/−×BALB/c ) F1 mice exhibited significantly higher 7 dpi infectious viremia compared to ( B6 mA3+/+×BALB/c ) F1 mice at all inoculum dosages , except at the lowest dose ( 14 SFFU ) , in which most titers were below the detection limit of the assay . The fold-difference in infectious viremia between ( B6 mA3+/+×BALB/c ) F1 and ( B6 mA3−/−×BALB/c ) F1 cohorts varied with the dose of the viral inoculum . The fold-difference in B6 mA3 restriction could not be accurately determined at 14 and 50 SFFU since several samples had infectious titers that were below the limit of detection . Maximum fold-difference in B6 mA3-mediated restriction was observed at 140 SFFU with 10-fold restriction ( Figure 3A ) . At 500 SFFU and 1400 SFFU inoculum doses , there was only a 4-fold and 2-fold effect , respectively ( Figure 3A ) . These results demonstrate that B6 mA3-mediated restriction was saturable in vivo . Notably , even at a higher viral dose that resulted in decreased B6 mA3 inhibition , similar viral RNA loads ( Figure 3B; left panel ) and higher noninfectious particle release ( Figure 3B; right panel ) were still observed in ( B6 mA3+/+×BALB/c ) F1 versus ( B6 mA3−/−×BALB/c ) F1 mice . The B6 mA3 gene promoted NAb responses in ( B6×A . BY ) F1 , ( B6×A/WySn ) F1 and pure B6 mice [20] , [27] . However , measurements of NAb responses at 28 dpi in ( B6×BALB/c ) F1 mice were confounded by the increased mortality of ( B6 mA3−/−×BALB/c ) F1 mice [20] , [26] . In contrast , infection with 10-fold lower dose ( 14 SFFU ) resulted in all of the mice surviving to 28 dpi , providing an opportunity to revisit this question in this F1 strain ( Figure 4A ) . As shown in Figure 4B , ( B6 mA3+/+×BALB/c ) F1 mice developed significantly stronger NAb responses than ( B6 mA3−/−×BALB/c ) F1 mice . Thus , the B6 mA3 gene promoted FV-specific NAb responses in four genetic backgrounds that include ( B6×BALB/c ) F1 , ( B6×A . BY ) F1 , ( B6×A/WySn ) F1 and B6 mice . To determine which components of the 28 dpi plasma correlated with B6 mA3-dependent neutralization , we evaluated the binding titers of the major immunologlobulin isotypes in plasma , IgM and IgG . Native virions were bound to 96-well plates , and endpoint IgM and IgG titers were determined by indirect ELISA . Endpoint IgM titers of 28 dpi plasma from B6 mA3+ F1 and B6 mA3− F1 were not significantly different from each other ( Figure S4 in Text S1 ) . In contrast , endpoint IgG titers against native virions were significantly higher in ( B6 mA3+/+×BALB/c ) F1 compared to ( B6 mA3−/−×BALB/c ) F1 mice ( Figure 4C , left panel ) . Notably , this difference was not detected if detergent lysed-virions were used ( Figure 4C , right panel ) . Similar results were observed for IgG endpoint titers in ( B6 mA3+/+×A . BY ) F1 versus ( B6 mA3−/−×A . BY ) F1 mice ( Figure S5 in Text S1 ) . Thus , the B6 mA3-dependent antibody response was distinguished by an IgG response directed against intact virus particles .
The molecular identification of the classical resistance gene Rfv3 as mA3 solved a 30-year mystery in retrovirology [20] , [26]–[27] . However , this discovery unlocked new questions , foremost of which is the mechanism for how mA3 promotes NAb responses . Recent studies suggested an indirect mechanism that linked the ability of B6 mA3 to restrict FV in vivo with a more vigorous B cell response [20]–[21] , [27]–[28] . However , this finding seemed counterintuitive in light of studies on other viral infections , particularly HIV-1 , which showed that viral antigen levels had to be preserved to maintain virus-specific antibody levels [45]–[47] . FV plasma viremia is routinely quantified using a focal infectivity assay that measures infectious virus [20] , [39] . We therefore investigated whether this method underestimated the total numbers of FV particles in mice with functional Apobec3 activity . Using a quantitative PCR assay for FV , B6 mA3+ F1 mice exhibited plasma viral RNA loads that were similar to B6 mA3-deficient F1 mice at 7 dpi . In other words , B6 mA3 activity led to no significant change in the physical numbers of virus particles . Instead , B6 mA3 activity reduced infectious virus titers , indicating the release of substantial levels of noninfectious FV particles . These B6 mA3-restricted plasma virions , which account for up to 80% of virions released during acute infection relative to B6 mA3− F1 mice , encounter a significant post-entry block in early reverse transcription in the next target cell , resulting in reduced FV infection in multiple cellular targets in vivo , including splenic B cells . Thus , high levels of B6 mA3-restricted FV particles likely drove the FV-specific B cell response that resulted in the development of potent NAbs ( Figure 5 ) . Native virions are potent inducers of humoral immunity . Repeating molecular patterns on virions may be particularly effective in cross-linking and activating B cell receptors , while viral nucleic acids could enhance B cell responses by activating Toll-like receptors [48] . The FV envelope glycoprotein is the primary target of NAbs [49]–[50] , and is organized as a trimeric spike in the native virion , analogous to the HIV-1 envelope glycoprotein [51]–[52] . Detergent treatment disrupts retroviral trimers into monomeric subunits . In this study , we show that the more potent NAb response from B6 mA3+ F1 mice at 28 dpi correlated with significantly higher IgG binding titers against native , but not detergent-lysed virions . Thus , B6 mA3-restricted virions primed a more effective IgG response directed against native envelope trimers . Further studies are in progress to characterize the molecular attributes of this protective NAb response . APOBEC3 is unique among known virus restriction factors due the circuitous nature of its inhibitory mechanism . Instead of simply restricting retroviruses in the infected cell , APOBEC3 evolved the ability to incorporate into budding virions and restrict intact virions in the next target cell [4]–[5] . This biological property of APOBEC3 is conserved from rodents ( mA3 ) to humans ( hA3G ) , suggesting an important evolutionary advantage . However , despite nearly 10 years since the discovery of this fundamental APOBEC3 phenotype [4]–[5] , the benefits of next-round inhibition by APOBEC3 to the host remain mysterious . Our findings suggest that mA3 restriction functions as an innate mechanism that allow B cell epitopes to be presented in the context of native virions , subsequently driving the NAb response . This humoral immune response is characterized by high specificity and memory , attributes that could allow the host to effectively control the infection as well as prevent subsequent infections . Although FV and HIV-1 infect different cell types and cause different diseases , functional similarities in APOBEC3 proteins from mice and humans suggest that concepts developed from the FV model may prove relevant to HIV-1 infection . Our current model on how mA3 promotes NAb responses implicates noninfectious virions as drivers of the Germinal Center ( GC ) cell response ( Figure 5 ) . These mA3 restricted virions retain immunogenicity but encounter a post-entry block in target cells that include B cells , reducing FV-induced immune dysfunction . While HIV-1 does not infect B cells , CD4+ T cells , the primary targets of HIV-1 , also perform critical functions in the GC response , directly interacting with B cells to promote antigen-specific antibody development [53] . Thus , augmenting hA3G function during acute HIV-1 infection may preserve CD4+ T cell function in GCs and promote HIV-specific antibody development . In addition , the role of noninfectious virions in driving the mA3/Rfv3 phenotype further support the use of native virion mimics such as virus-like particles or stabilized trimers as base scaffolds for vaccine design [54]–[55] , with the caveat that more sophisticated approaches are needed to elicit NAbs that could broadly neutralize HIV-1 strains from multiple subtypes . The existence of a lentiviral A3G antagonist , Vif , provides an opportunity to experimentally test whether modulating A3G function can improve the lentivirus-specific humoral immune response . In the SIV model , mutating the Vif gene in SIV to attenuate its function [56] may allow for rhA3G to promote non-infectious virion release and improve humoral immunity in infected rhesus macaques . Similar studies on HIV-1 infection in humans are not possible , but therapeutic agents that block the Vif-hA3G interaction could prove useful . Unfortunately , compounds independently confirmed to specifically inhibit the Vif-hA3G interaction have yet to be identified . In this study , we provide the first evidence that innate mA3 restriction is saturable in vivo , possibly reflecting a delicate balance between the endogenous levels of mA3 and a putative mA3 antagonist encoded by FV , Glyco-Gag [57] . Thus , inducing hA3G levels to saturate , rather than disrupt , the interaction with Vif may be a viable alternative to promote hA3G activity ( Figure 5 ) . Notably , Interferon-alpha ( IFN-α , a cytokine that could induce APOBEC3G expression in HIV-1 target cells in vitro [58]–[60] , improved the kinetics of HIV-1 specific antibody development when clinically administered during acute HIV-1 infection in vivo [61] . Further studies on the link between IFN-α and APOBEC3 in vivo may provide critical insights on whether the saturability of APOBEC3 restriction can be exploited for therapy and vaccine development .
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 University of Colorado Health Sciences Center Animal Care and Use Committee [Permit Number B-89709 ( 10 ) 1E] . All infections were performed under isoflurane anaesthesia , and all efforts were made to minimize suffering . B6 , BALB/c and A . BY mice were purchased from The Jackson Laboratory . B6 mA3 deficient mice were derived from the XN450 cell line [20] and backcrossed for 9 generations . Experimental groups consist primarily of ( B6 mA3+/+×BALB/c ) F1 versus ( B6 mA3−/−×BALB/c ) F1 mice . The rationale for an F1 transcomplementation approach is explained in more detail ( Figure S1 in Text S1 ) . Experiments were also performed in ( B6 mA3+/+×A . BY ) F1 versus ( B6 mA3−/−×A . BY ) F1 mice ( Figures S2 , S4 and S5 in Text S1 ) . Note that B6 , BALB/c and A . BY mice have a functional B cell Activating Factor Receptor ( BAFF-R ) and normal B cell maturation phenotype , as previously described [26] . These mice are also Fv1b and are therefore susceptible to B-tropic FV infection [62] . Mice were infected with FV complex derived from in vivo-passaged stocks originally used to describe Rfv3 [24] , [63] . This B-tropic FV stock contains: replication-competent ecotropic Friend murine leukemia helper virus ( F-MuLV ) ; replication-defective spleen-focus forming virus ( SFFV; Lilly-Steeves strain ) ; lactate-dehydrogenase elevating virus ( LDV ) . LDV is a ‘contaminant’ RNA virus that could enhance the pathogenicity of FV by delaying adaptive immune responses [64]–[65] . No polytropic or mink-cell focus-inducing viruses ( MCFs ) were detectable in the virus stock by focal immunofluorescence assay with antibodies Hy7 or mAb 516 , which detect the vast majority of MCFs [66] ( detection limit of 20/ml ) . SFFV titers in the FV stock were titered in BALB/c mice and expressed as spleen focus forming units ( SFFU ) per ml . ( B6×BALB/c ) F1 mice were infected intravenously via the retro-orbital route with 14 to 1400 SFFU ( spleen focus forming units ) in 300 µl RPMI and ( B6×A . BY ) F1 mice were infected with 1400 SFFU . All mice were >2 months old . Plasma samples were harvested at 7 or 28 days post-infection ( dpi ) . Infectious viremia titers were determined by serially diluting plasma into Mus dunni cells containing 4 µg/ml polybrene ( Sigma; St Louis , MO ) . Infected cells were detected using a monoclonal antibody specific to F-MuLV gp70 , MAb 720 , as previously described [20] , [39] . Briefly , F-MuLV gp70+ cells were detected following incubation with anti-mouse IgG conjugated to horseradish peroxidase and an insoluble substrate , 3-amino-9-ethylcarbazole ( Sigma ) . Titers were expressed as log10 focus forming units ( FFU ) per ml of plasma . Viral RNA copy numbers were quantified by real-time PCR ( qPCR ) as described [26] , [40] . RNA from plasma ( 10 µl ) was extracted using the RNAeasy kit ( Qiagen; Valencia , CA ) , and used as template for one-step reverse transcription and PCR ( Applied Biosystems; Carlsbad , CA ) using FV-specific primers ( FLVsense: 5′-GGACAGAAACTACCGCCCTG and FLVantisense: 5′-ACAACCTCAGACAACGAAGTAAGA ) and probe ( FLVprobe: FAM-TCGCCACCCAGCAGTTTCAGCAGC-TAMRA ) . Copy numbers were interpolated from an in-plate T7-transcribed RNA standard , and expressed as log10 copy number per ml of plasma . Sequence alignments of the F-MuLV helper virus ( GenBank Accession #Z11128 ) , the Lilly-Steeves SFFV strain ( V01552 . 1 ) and MCFs ( L . H . Evans , unpublished data ) revealed that the FV qPCR primers have low to no significant identity with SFFV or MCFs ( data not shown ) . Furthermore , we tested whether the FV qPCR primers could detect any endogenous polytropic env sequences present at one copy per mouse genome . The FV qPCR assay consistently detects an input of 100 copies of cloned F-MuLV DNA . If the qPCR primers cross-react with endogenous MLV , then we should obtain a positive signal if >100 copies of uninfected genomic DNA are added into the reaction . An input of 100 ng genomic DNA ( ∼34 , 000 genomes ) from uninfected B6 , A . BY and BALB/c mice into the qPCR reaction yielded no detectable signals . In contrast , >105 copies were detected from 100 ng genomic DNA from FV infected B6 mice at 7 dpi ( S . X . Li and M . L . Santiago , unpublished ) . Thus , we conclude that the qPCR assay is specific for the F-MuLV helper virus . RNA was extracted from the FV inoculum and acute plasma using the RNAEasy kit ( Qiagen ) and reverse-transcribed using the RT2 cDNA synthesis kit ( SA Biosciences; Frederick , MD ) . FV env sequences ( 849 bp ) were obtained by amplifying with primers FV . f ( ACTTATTCCAACCATACCTCT ) and FV . r ( TTTAGCTGGTGGTATTGTTGA ) using the Phusion Hi-Fidelity PCR kit ( Finnzymes; Woburn , MA ) . Amplicons were cloned using the TOPO cloning kit ( Invitrogen; Carlsbad , CA ) . FV inoculum sequences were aligned with FB29 and PVC-211 ( GenBank Z11128 and M93134 ) using ClustalX ( http://www . clustal . org/ ) . Phylogenetic trees were constructed using the neighbor joining method with 1000 subreplicates . PVC-111 was used as outgroup . Viral RNA sequences were compared with the FV inoculum consensus , excluding variations that were already detected in the inoculum . To quantify mutational loads , the total , G-to-A and mA3-associated mutational frequencies relative to the consensus were divided by the total number of base pairs , G nucleotides and GG/GA dinucleotides analyzed , respectively . To bias the detection of G-to-A mutations , we combined four approaches [11] , [41]: ( 1 ) reverse transcripts were amplified following a single-round infection of Mus dunni cells , to enrich for potentially defective reverse transcripts; ( 2 ) a segment of env closest to the primer binding site was chosen , since this region may be present in single-stranded DNA form for longer duration and thereby more susceptible to deamination [67]; ( 3 ) Taq polymerase was utilized instead of Pfu polymerase , since Pfu polymerase activity is inhibited by deoxyuridines in DNA templates; and ( 4 ) a denaturation temperature of 88°C was used to enrich the detection of G-to-A hypermutated reverse transcripts , which should have a lower melting temperature . 7 dpi plasma samples ( 2 . 5 µl ) from wild-type and B6 mA3− F1 mice were inoculated into Mus dunni cells in a 6-well plate containing 4 µg/ml polybrene . DNA was extracted after 2 days using the DNAeasy kit ( Qiagen ) . PCR was performed with 10 ng DNA , 1× Sweet PCR mix ( SA Biosciences ) , 2 . 5 mM dNTP , 1 . 25 pmol env primers FV . f and FV . r . Thermocycling conditions included a 95°C 15 min hot-start , followed by 30 cycles of denaturation at 88°C for 30 s , 55°C for 30 s and 72°C for 1 . 5 min . Amplicons were cloned using the TOPO-TA cloning kit ( Invitrogen ) and 4–8 clones were sequenced for each sample . Consensus from the FV inoculum and the corresponding plasma viral RNA sequences were used as reference for analysis of nucleotide substitutions using the HYPERMUT 2 . 0 software ( hiv . lanl . gov ) . DNA samples as described above were subjected to absolute quantifications for early ( R-U5 ) and late ( R-gag ) FV transcripts , as well as a housekeeping gene , beta-actin , using a Taqman assay ( Applied Biosystems; Foster City , CA ) in a CFX96 real-time system ( Bio-Rad; Hercules , CA ) . The primers and probes are listed as follows . Early reverse transcripts: R-U5 . fwd , CTCCGATAGACTGAGTCG , R-U5 . rev , AGACCCTCCCAAGGAACA , R-U5 . probe FAM-CCCGTGTATCCAATAAATCCTCTTGC-TAMRA . PCR cycling conditions were 95°C 10 min followed by 40 cycles of 95°C for 15 sec and 55 . 7°C for 45 sec . The expected size of PCR product is 94 bp . Late reverse transcripts: R-U5 . fwd and R-Gag . rev , TTCGACATCCTTCCAGTGGT and R-Gag . probe , FAM-CTGCAGCATCGTTCTGTGT-TAMRA . The PCR conditions for R-Gag were 95°C for 10 min followed by 40 cycles of 95°C for 15 sec and 61°C for 2 min 30 sec . The expected size of the R-Gag amplicon is 670 bp . Mouse beta-actin: Actin . fwd , GGCACCACACCTTCTACAATG , Actin . rev , GGGGTGTTGAAGGTCTCAAAC , and Actin . probe FAM-TGTGGCCCCTGAGGAGCACCC-TAMRA . PCR conditions included a hot-start for 95°C 10 min followed by 40 cycles of 95°C for 15 sec and 60°C for 50 sec . Absolute copy numbers were interpolated from a best-fit standard curve against a DNA standard . Bone marrow and spleen cells ( 106 cells ) were stained with MAb 34 , an IgG2b monoclonal antibody specific for FV Glyco-Gag [44] for 30 min , then co-stained with: Ter119-FITC ( clone TER-119 ) , CD3-Alexa700 ( 17A2 ) , ( BD Biosciences; San Diego , CA ) ; CD11c-PE-Cy7 ( N418 ) , ( eBioscience; San Diego , CA ) ; CD19-PerCP-Cy5 . 5 ( 6D5 ) ( Biolegend; San Diego , CA ) and anti-mouse IgG2b-APC ( Columbia Biosciences; Columbia , MD ) . Isotype controls and cells from uninfected mice were used for gating . Cells were processed in an LSR-II flow cytometer ( BD Biosciences ) , collecting up to 250 , 000 events per sample . Datasets were analyzed using the Flowjo software ( Treestar; Ashland , OR ) . Serial fourfold dilutions of heat-inactivated plasma were incubated with F-MuLV-N stock virus in the presence of guinea pig complement ( Sigma ) . The antibody∶virus mixture was added into M . dunni cells and developed as in the plasma virus titrations [20] . The number of colonies were counted and compared to a no antibody control , which was set as 100% . Neutralization curves were constructed , and IC50 values were calculated based on a one-site sigmoidal fit using the Graphpad Prism software ( Irvine , CA ) . Virion antigens were prepared from culture supernatants obtained from Mus dunni cells infected with F-MuLV-N in T-175 flasks in the presence of 4 µg/ml polybrene ( Sigma ) . Cellular debris was pelleted at 1800× g for 5 min at 4°C , then the supernatants were passed through a 0 . 22 µm filter . The filtered supernatants were ultracentrifuged at 25 , 000× g for 2 hr at 4°C in SW28 Ultra-Clear tubes ( Beckman Coulter; Brea , CA ) , and virion pellets were resuspended in 0 . 5 ml Tris Buffered Saline ( TBS ) containing 1× protease inhibitor cocktail ( Calbiochem; La Jolla , CA ) per tube and allowed to dissociate overnight at 4°C . In some preparations , the virion pellets were resuspended in TBS with 1% Empigen-BB ( Sigma ) , a zwitterionic detergent that is commonly used to solubilize viral particles and liberate envelope monomers [68] , [69] . Virions were aliquoted and total protein concentrations were determined using a BCA protein assay ( Pierce; Rockford , IL ) . ELISAs were performed at room temperature and 100 µl/well volumes unless otherwise indicated . Virions ( 200 ng per well ) were coated into Immulon-4 HBX plates ( Thermo Scientific Nunc; Rochester , NY ) overnight at 4°C and blocked with SuperBlock ( Pierce; Rockford , IL ) for 2 hr . Serial 2-fold dilutions of plasma in Phosphate Buffered Saline ( PBS ) were added and incubated for 1 hr . After 6 washes with PBS with 0 . 05% Tween-20 ( PBS-T ) , 1∶4000 biotinylated goat anti-mouse IgG ( Southern Biotechnology; Birmingham , AL ) was added and incubated for 1 hr . After 6 PBS-T washes , 1∶4000 streptavidin-conjugated horseradish peroxidase ( Southern Biotechnology ) was added and incubated for 30 min . Following 6 PBS-T washes , 100 µl of TMB substrate ( BioFX Laboratories; Owings Mills , MD ) was added per well and incubated in the dark for 15 min . The reaction was stopped with 0 . 3N sulfuric acid ( Sigma ) , and absorbances were read at 450 nm in a Victor X5 plate reader ( Perkin Elmer; Waltham , MA ) . The same procedures were followed for the IgM ELISAs , except that plasma samples were incubated overnight at 4°C , and 1∶10000 biotinylated goat anti-mouse IgM ( Southern Biotechnology ) was used . Endpoint titers were calculated by constructing one-site total best-fit curves using the Graphpad Prism software , and interpolating plasma concentrations that correspond to a cut-off value based on twice the mean absorbance background from wells with no plasma added . Samples from uninfected mice were also used as negative controls , and had absorbance values below the cut-off value ( data not shown ) . Endpoint titers were expressed as log2 concentrations . Differences between means were analyzed using a two-tailed Student's t test . The association between G-to-A mutations with B6 mA3 status was inferred by subjecting 2×2 contigency tables to a two-tailed Chi-square test . Differences with p values >0 . 05 were considered not statistically significant ( N . S . ) .
|
Members of the APOBEC3 gene family can potently inhibit a broad range of retroviruses , including HIV-1 . In cell culture , APOBEC3 counteracts retroviruses by: ( 1 ) reducing the infectivity of virions; and ( 2 ) inducing lethal G-to-A hypermutation in the next target cell . The selective advantage to the host of an ‘indirect’ restriction factor that is incorporated into virions and acts in the next target cell remains mysterious . We previously showed that Apobec3 encodes Rfv3 , a classical resistance gene that controls the neutralizing antibody response against Friend retrovirus infection in mice . Here we demonstrate that Apobec3 promotes the release of substantial levels of noninfectious virions in the plasma during acute FV infection , resulting in a more potent antibody response directed against intact virions . Thus , we propose that APOBEC3 evolved as an innate mechanism to promote high concentrations of retrovirus antigen in a native but noninfectious form to effectively prime the neutralizing antibody response . These findings could have important implications for improving HIV-1 specific antibody responses .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"humoral",
"immunity",
"animal",
"models",
"of",
"infection",
"viral",
"vaccines",
"mechanisms",
"of",
"resistance",
"and",
"susceptibility",
"immunity",
"virology",
"innate",
"immunity",
"immune",
"defense",
"biology",
"microbiology",
"viral",
"evolution",
"pathogenesis"
] |
2011
|
Noninfectious Retrovirus Particles Drive the Apobec3/Rfv3 Dependent Neutralizing Antibody Response
|
Endosymbiotic Wolbachia bacteria are potent modulators of pathogen infection and transmission in multiple naturally and artificially infected insect species , including important vectors of human pathogens . Anopheles mosquitoes are naturally uninfected with Wolbachia , and stable artificial infections have not yet succeeded in this genus . Recent techniques have enabled establishment of somatic Wolbachia infections in Anopheles . Here , we characterize somatic infections of two diverse Wolbachia strains ( wMelPop and wAlbB ) in Anopheles gambiae , the major vector of human malaria . After infection , wMelPop disseminates widely in the mosquito , infecting the fat body , head , sensory organs and other tissues but is notably absent from the midgut and ovaries . Wolbachia initially induces the mosquito immune system , coincident with initial clearing of the infection , but then suppresses expression of immune genes , coincident with Wolbachia replication in the mosquito . Both wMelPop and wAlbB significantly inhibit Plasmodium falciparum oocyst levels in the mosquito midgut . Although not virulent in non-bloodfed mosquitoes , wMelPop exhibits a novel phenotype and is extremely virulent for approximately 12–24 hours post-bloodmeal , after which surviving mosquitoes exhibit similar mortality trajectories to control mosquitoes . The data suggest that if stable transinfections act in a similar manner to somatic infections , Wolbachia could potentially be used as part of a strategy to control the Anopheles mosquitoes that transmit malaria .
Bacterial associates are ubiquitous among insects , including mosquitoes [1] . Wolbachia are obligate endosymbiotic bacteria that infect numerous insects , many of which are vectors of pathogenic microorganisms . Much interest has centered around Wolbachia as a means of reducing arthropod-borne disease due to the capacity of the bacteria to manipulate the reproduction of the insect host , which in turn favors their own transmission [2] , [3] . However , recent studies detail that Wolbachia can directly cause pathogen interference ( PI ) in their invertebrate hosts , whereby infected insects are less susceptible to pathogens [4] , [5] , [6] , [7] , [8] , [9] . Fitness benefits conferred by PI may partially explain the prevalence of Wolbachia strains that do not confer the more familiarly known reproductive manipulations such as cytoplasmic incompatibility . For example , some Drosophila species infected with specific Wolbachia strains have greater resistance to viral pathogens compared to their uninfected counterparts [4] , [9] , [10] . From an applied standpoint , mosquito vectors artificially transinfected with Wolbachia exhibit PI against diverse pathogens [5] , [6] , [8] . The heterologous association between Wolbachia and novel host seems to strongly induce this phenotype in mosquitoes , as the native Wolbachia strain in many vectors does not generally affect pathogen transmission [6] , [8] . Wolbachia does cause a small reduction in West Nile virus titer in Culex quinquefasciatus , but this effect is subtle and is unlikely to affect the vector competence of the mosquito [7] . In Aedes aegypti , artificial Wolbachia infections suppress diverse pathogens including RNA viruses , filarial nematodes and the avian malaria parasite Plasmodium gallinaceum [5] , [6] , [8] . In Anopheles mosquitoes , somatic infection with the Wolbachia strain wMelPop suppresses the rodent malaria parasite P . berghei . These results show that Wolbachia-induced PI may be of use to control various vector-borne diseases [11] . Although the mechanism behind Wolbachia-induced PI is uncertain , several non-mutually exclusive hypotheses have been proposed . In wMelPop and wAlbB-transinfected Ae . aegypti , there is induction of the basal immune state of the host by the novel Wolbachia strain [5] , [6] , [8] . Activation of the immune state before the mosquito is challenged with pathogens may make the insect less susceptible to infection . Additionally , there is evidence for resource competition between Wolbachia and pathogens such as dengue virus , where virus was only observed in mosquito cells that were not infected with Wolbachia [6] . In addition to PI and manipulation of host reproduction , the wMelPop strain of Wolbachia causes life shortening in both Drosophila and transinfected Aedes aegypti [12] , [13] . Due to the extrinsic incubation period ( EIP ) of many pathogens , life shortening can have a dramatic effect on reducing pathogen transmission . As such , wMelPop has been proposed to control vector-borne diseases by skewing the age structure of the mosquito population toward the younger age classes that are not old enough to transmit pathogens [14] , [15] . The dual effect of life shortening and PI can act synergistically , enhancing the prospects for Wolbachia-based disease control strategies [5] , [6] , [12] . Although naturally uninfected , Anopheles mosquitoes are amenable to Wolbachia infection , both in vitro [16] and in the mosquito somatic tissues [17] . Somatic infection of insects allows for evaluation of Wolbachia phenotypes in the absence of a stably infected host . Recently , somatic infection by wMelPop in An . gambiae was shown to reduce P . berghei levels in conjunction with induction of several innate immune genes . However , immune up-regulation was only investigated at a single time point [11] . It is unknown whether immune induction occurs constantly throughout the life of the mosquito , whether Wolbachia infection will modulate Plasmodium species that are important for human health concerns , or whether different Wolbachia strains will induce similar phenotypes . To address these issues , we characterized the infection dynamics of two divergent Wolbachia strains ( wMelPop and wAlbB ) in somatically infected An . gambiae , using fluorescence in situ hybridization ( FISH ) and qPCR . Host immune gene expression in response to Wolbachia infection was assessed at multiple time points throughout the lifespan of the mosquito . Wolbachia mediated PI was evaluated for the human pathogen P . falciparum . We show that the mosquito immune response to Wolbachia is dynamic , switching between induction and suppression as the mosquito ages . We examined life history traits of mosquitoes infected with the life shortening strain of Wolbachia wMelPop , before and after bloodmeals , and show that strong life shortening was only observed immediately after bloodfeeding . The results are discussed in terms of potential applications for using Wolbachia as part of a strategy for malaria control .
Using whole mosquito fluorescence in situ hybridization ( FISH ) , we determined that the Wolbachia strain wMelPop disseminates throughout the mosquito and infects numerous tissues after somatic infection by thoracic microinjection . By 30 days post-infection , Wolbachia is ubiquitous in the abdomen , where it primarily resides within cells of the fat body , and in cells that adhere to the Malpighian tubules , which are most likely hemocytes that have phagocytized Wolbachia . The fat body and hemocytes are major immune tissues within the mosquitoes and infection of these tissues could potentially affect immune processes . Previously it had been demonstrated that Wolbachia could replicate within Anopheles mosquitoes , however the cellular orientation of the infection was unknown [17] . The occurrence of Wolbachia within fat body and hemocyte cells demonstrate conclusively that Wolbachia have the capacity to enter , replicate and survive intracellularly in specific somatic tissues within Anopheles . This observation is supported by in vitro experimentation where Wolbachia has established infections in Anopheles cell culture [16] . Wolbachia are also observed to infect the head of the insect , possibly in the brain or pericerebral fat body . Infection is also observed within the mouthparts and sensory organs of the mosquito ( Figure 1 ) – whether these Wolbachia are free in the hemolymph or contained within circulating hemocytes remains to be determined . The distribution of Wolbachia in somatically infected An . gambiae in part resembles that of the stably infected Aedes aegypti [6] , [12] . One noticeable difference between the two mosquito species is the lack of infection in the Anopheles midgut and germline ( Figure S1 ) . Although adult microinjection has successfully been adapted to transinfect multiple insect species [18] , [19] , [20] , no evidence was found for entry of wMelPop into the An . gambiae germline . Previously , adult injection was successfully used to re-infect D . melanogaster with wMel , and to establish infection in Ae . aegypti with wAlbA and wAlbB [18] , [20] . Laodelphax striatellus , which naturally harbors wStri , was co-infected with wRi using adult microinjection [19] , while wStri has been transferred to Nilaparvata lugens by nymphal injection [21] . In D . melanogaster , Wolbachia was localized to the somatic stem cell niche in the germarium [20] , while in both Ae . aegypti and L . striatellus , progeny of microinjected females were infected suggesting entry of Wolbachia into the germline [18] , [19] . In contrast , and similar to our results , somatic infection of Bombyx mori was successful after microinjection of Wolbachia into immature life stages , but germline infection was not established [22] . Using FISH , no signal was detected in mature ovaries or immature ovarioles in Anopheles ( Figure S1 ) . The lack of infection of the An . gambiae germline may go some way to explain the unique biology of the Anopheles genera , which is naturally uninfected in nature and seems to be impervious to Wolbachia transinfection despite numerous attempts . There are many possibilities that may explain the lack of infection in the ovary . While Wolbachia can survive intracellularly in Anopheles mosquitoes , the ovarian milieu may be inhospitable to the bacteria . Alternatively , ovarian cell receptors that Wolbachia utilizes may be too divergent in Anopheles , preventing entry into the ovary . Infection itself may cause reproductive ablation . Amhed and Hurd [23] demonstrated that apoptosis in ovarian follicular epithelial cells occurs when the melanization response or humoral antimicrobial activity is induced in An . gambiae . Alternatively , constraints to infection may be related to the bacteria . It is evident that Wolbachia can adapt to new host backgrounds [24] , and certain strains of Wolbachia may be more or less suitable for infection establishment . Experiments that address these hypotheses may provide a mechanistic basis for the inability of Wolbachia to infect the Anopheles germline and may provide clues that could ultimately lead to transinfection of this genus . Quantitative PCR ( qPCR ) analysis demonstrated that Wolbachia multiples within the mosquito . Since we do not know whether Wolbachia are polyploid , results are presented as Wolbachia genomes per host genome . After microinjection , there is an initial decrease in bacterial density before Wolbachia replicates to increase in abundance ( Figure 2 ) . These results are in concordance with Jin et al [17] who used standard PCR to assess somatic infection dynamics of the wMelpop Wolbachia strain . Here , we quantify both wMelpop and wAlbB infection with qPCR and find both these Wolbachia strains display a similar infection pattern , although wAlbB densities are several orders of magnitude lower than wMelPop . This is not unexpected as wMelPop , an over replicating strain , replicates faster than wAlbB in the mosquito ( Figure 2 ) and is initially extracted from cell culture and microinjected into the mosquito at higher densities . It is also possible that the ploidy of wMelPop is higher than wAlbB . In contrast to Ae . aegypti stably infected with Wolbachia , we see that the immune response in Anopheles after somatic infection is dynamic . At 3 days post infection there is minimal effect on gene expression . Infection by wMelPop and wAlbB moderately suppress Serpin6 . wMelPop moderately suppresses cactus , the negative regulator of the Toll pathway , while wAlbB moderately induces Caspar , the negative regulator of the IMD pathway . At 6 days post-infection , Caspar is suppressed by wMelPop in conjunction with up-regulation of Rel2 and cecropin , as well as modestly up-regulating cactus . This time period is coincident with the initial clearing of infection measured by qPCR ( Figure 2 ) , and is similar to observations by Kambris and colleagues [11] who observed immune up-regulation ( including strong cecropin induction ) at a similar time point ( 8 days post-infection ) . wAlbB infected mosquitoes display a different profile at this time point , with gene expression not significantly affected . However , at 10 post-infection , the pattern changes to dramatic down-regulation of many immune-related host genes in response to both Wolbachia strains , including FBN9 , Heat shock 70 , CLIP7A , TEP15 and the transcription factors Rel1 and Rel2 ( Figure 3 ) . This time period corresponds with Wolbachia replication in the mosquito ( Figure 2 ) , suggesting that Wolbachia may be actively manipulating host gene expression to mediate the infection and replication process . In several instances , suppression of host gene expression by wAlbB is greater compared to wMelPop , suggesting there are strain-specific responses in addition to differences related to bacterial density . This down-regulation is in agreement with regulation patterns observed in vitro , where the Wolbachia strains wAlbB and wRi suppressed many host genes ( including genes associated with innate immunity ) in cultured An . gambiae Sua5B cells [25] . By 15 days post infection , the response is mixed , with some genes up-regulated and some down-regulated in a Wolbachia strain-specific manner ( Figure 3 ) . After somatic infection , P . falciparum oocyst development was significantly reduced ( 40–60% ) by both wMelPop and wAlbB compared to the Mos55 ( Anopheles cell extract ) injected control . We observed similar results using both low gametocytemic and high gametocytemic Plasmodium cultures ( Figure 4 ) . In the low gametocytemic replicate , infection prevalence ( percentage of mosquitoes with one or more oocysts per midgut ) was statistically reduced in wMelPop-injected mosquitoes ( Mos55: 75% , N = 65; wMelPop: 33% , N = 21; wAlbB: 60% , N = 45; d . f . = 2 , Cramer's V = 0 . 39 , P = 0 . 002 ) . Infection prevalence did not differ statistically in the high gametocytemic replicates ( Mos55: 90% , N = 50; wMelPop: 83% , N = 35; wAlbB: 84% , N = 55 ) . No correlation was observed between Wolbachia density and Plasmodium oocyst load for either Wolbachia strain ( Figure S2 ) , suggesting that the reduction of Plasmodium is not directly related to Wolbachia density ( i . e . mosquitoes with high oocyst levels did not necessarily have the lowest Wolbachia titers ) . While wMelPop moderately induces the mosquito immune system at 6 days post-infection , by 10 days post-injection , the majority of tested immune genes were down-regulated by both Wolbachia strains ( Figure 3 ) . These time points correlate to when Plasmodium is developing within the mosquito midgut . Although , Kambris et al [11] provide evidence that wMelPop-mediated immune up-regulation induces PI in Anopheles against P . berghei , our data suggest that the mosquito immune response to Wolbachia is more dynamic . The modulation of the later immune response suggests mechanisms other than stimulation of basal immunity may be involved in PI in An . gambiae . Alternatively , immune up-regulation around the initial infection period when ookinetes are invading the midgut may be sufficient for a decrease in Plasmodium load . Possibly these different mechanisms may be acting in concert . A more thorough analysis of global immune regulation in response to Wolbachia infection throughout the life of the insect may clarify this issue . In our Plasmodium experiments , we noted higher mortality of wMelPop-injected mosquitoes compared to wAlbB or cell homogenate-injected treatments . Our previous data suggested that somatic infections of wMelPop were not virulent to Anopheles gambiae [17] . However , in those experiments mosquitoes were not allowed access to blood . We therefore considered the hypothesis that wMelPop-induced virulence in Anopheles gambiae was conditional on bloodfeeding . Mosquitoes were injected with wMelPop or with uninfected cell culture homogenate as previously described , held for 7 days , then were offered a human bloodmeal with or without P . falciparum parasites through a membrane feeder . After bloodfeeding , fed mosquitoes were separated from unfed mosquitoes and their mortality trajectories assessed . We observed that prior to bloodfeeding , there were no dramatic differences in mortality between infected and uninfected mosquitoes , similar to previous observation . However , wMelPop-infected mosquitoes exhibited a dramatic increase in mortality between 12–24 h post-bloodmeal . After 3 days approximately 80% of the mosquitoes died . After this period , the mortality trajectories of the two treatments become similar again ( Figure 5 ) . Infection with Plasmodium made no difference in the mortality phenotypes . Interestingly , we also noted that when comparing Wolbachia levels to Plasmodium oocyst levels , Wolbachia titers were much lower in assayed wMelPop-infected mosquitoes compared to wAlbB mosquitoes ( Figure S2 ) , suggesting that mosquitoes with high wMelPop titers did not survive long enough to be assayed for Plasmodium infection . These data show that wMelPop is virulent to An . gambiae , but the virulence phenotype is different than that described for Ae . aegypti and Drosophila [12] , [13] . Instead of a general increase in lifetime mortality rates , we observe an acute increase in mortality directly related to bloodmeal acquisition and/or digestion . Post bloodmeal , multiple developmental and metabolic processes occur which drastically alter mosquito physiology . Alteration of any of these processes by Wolbachia may potentially induce mortality . In cultured Anopheles Sua5B cells , Wolbachia infection down-regulates host expression of multiple antioxidant genes , including peroxiredoxin , superoxide dismutase and glutathione S transferase [25] . In bloodfed mosquitoes , antioxidant transcripts are up-regulated post bloodmeal [26] , [27] , [28] , [29] , [30] . A blood meal also increases iron levels , which are a precursor to reactive oxygen species ( ROS ) . In other systems , Wolbachia has been seen to influence the expression of ferritin and plays a role in iron metabolism [31] , [32] . We hypothesize that modulated levels of ROS within the mosquito may be the cause of post bloodmeal mortality . Lending credence to this hypothesis is the observation of increased mortality post-bloodmeal in An . gambiae after silencing of anti-oxidant genes [33] . The more striking mortality observed in this study may be due to down-regulation of numerous genes . Additionally , blood feeding is known to spark a proliferation of bacteria within the insect [1] . In Ae . aegypti , the expansion of gut bacteria post blood meal is attributed to a reduction in ROS , which can result in death of the mosquito [34] . Here , pathogenicity may be directly linked to wMelPop levels or indirectly by Wolbachia influencing the density of other bacteria . Alternatively , the effect of wMelPop on other physiological processes that occur after a blood meal ( such as vitellogenesis or nutrient metabolism ) may cause fitness costs , as seen in Ae . aegypti where wMelPop affects reproductive output when mosquitoes were fed on non-human hosts [35] . If stable Anopheles infections behave in a similar manner to somatic infections , this acute mortality phenotype could inhibit CI-induced drive of wMelPop into mosquito populations , and provide a selection pressure against the life-shortening phenotype as a large proportion of mosquitoes may die before producing offspring . These potential pitfalls could be offset by the use of this phenotype in a population suppression strategy , or the use of non-virulent Wolbachia strains such as wAlbB . The use of Wolbachia to control arthropod-borne disease has been postulated for some time . Previous ideas centered on the use of Wolbachia as a gene drive agent , however now it is evident that Wolbachia can also inhibit pathogen development in insects [4] , [5] , [6] , [7] , [8] , [9] . The obvious limitation to this approach for malaria control is the failure to create a Wolbachia infected Anopheles line , and this still remains a massive challenge in the field of Wolbachia biology . Here we have shown that An . gambiae mosquitoes somatically infected by two strains ( wMelPop or wAlbB ) are less susceptible to the major human malaria parasite P . falciparum . Using FISH and qPCR , we determined that Wolbachia has ubiquitous distribution in many mosquito tissues and replicates within the Anopheles host . As one oocyst is capable of producing many sporozoites , it would be interesting to determine if sporozoite number is reduced by Wolbachia considering the vast tissue distribution in somatically infected mosquitoes . The results suggest that An . gambiae stably infected with Wolbachia may have reduced ability to maintain transmission of Plasmodium by multiple strain-dependent mechanisms .
Anonymous expired human blood was obtained from a local blood bank for use in mosquito blood feeding experiments . Wolbachia was cultured and extracted from infected Anopheles cells as previously described [16] , [36] . An . gambiae mosquitoes ( Keele strain ) were reared as described [16] . Two days post emergence , adult female mosquitoes were anesthetized on ice and injected with Wolbachia according to previously established methodology [17] . Post injection , mosquitoes were incubated at 19°C for 2 days for recovery then maintained at 28°C . FISH was performed on wMelPop infected mosquitoes 30 days post injection following the experimental procedure outlined by Koga et al . [37] . Briefly , mosquitoes were fixed in acetone for 3 months , legs were removed and mosquitoes were secondarily fixed in Carnoy's solution . To minimize autofluorescence , mosquitoes were transferred to 10% hydrogen peroxide in 6% alcohol for 5 days . After rehydration in PBST ( 1–2 hours ) , tissues were pre-hybridized followed by hybridization with the Wolbachia specific probe overnight [38] . Samples were washed in PBST 3 times to remove excess probe , counterstained with SYTOX green ( Invitrogen ) and visualized by epifluorescent and confocal microscopy . Individual channel images are available as Supplementary data ( Figure S3 ) . FISH controls included 1 ) no probe controls , 2 ) competition controls in which unlabeled oligonucleotides were added to the hybridization buffer to suppress the fluorescent signals and 3 ) RNase digestion controls , in which prior to hybridization RNAs in the insect materials were removed by RNase A treatment ( Figure S4 ) . DNA or RNA was extracted from somatically infected mosquitoes using DNAzol ( Molecular Research Center , Inc . , Cincinnati , OH ) or RNeasy mini kits ( Qiagen ) for estimation of Wolbachia density and quantification of host gene expression respectively . qPCR to determine the density of wMelPop in whole mosquitoes was completed by amplifying the single copy gene WD_0550 [24] , while wAlbB was amplified with modified GF and BR primers which specifically bind to the wsp gene [18] . Ten mosquitoes were assay at each time point for each strain to estimate Wolbachia density , while 5 mosquitoes were used for host gene expression per time point . The relative abundance of each Wolbachia strain was determined after normalization to the mosquito single-copy S7 gene [39] . For host gene expression , RNA was DNase treated ( Ambion ) and cDNA synthesized using superscript III ( Invitrogen ) following manufactures guidelines . qPCR was completed using a Rotor gene Q ( Qiagen ) using the Rotor gene SYBR green PCR kit ( Qiagen ) according to manufactures guidelines . qPCRs were completed in triplicate . PCR primers are listed in Table S1 . Melt curve analysis was completed on all PCRs . In Wolbachia density experiments , data were analyzed by Kruskal-Wallis test using the Connover-Inman method for pairwise contrasts between time points . For host gene expression experiments , significance was assessed by Mann-Whitney U test compared to mosquitoes injected with uninfected Mos55 cell culture homogenate ( control ) . Tested mosquito genes were identified in a microarray screen of Wolbachia-regulated Anopheles genes in cultured cells [25] . Additional analyses were conducted using REST [40] and qGENE [41] software . 2-day old female mosquitoes were intrathoracically injected with wMelPop or wAlbB ( purified from cell culture ) as described [17] or with uninfected Mos55 cell culture homogenate ( control ) . Seven days post-injection , mosquitoes were offered a Plasmodium-infected blood meal . Prior to blood feeding , mosquitoes were starved overnight . The gametocytemia of infected blood meals was approximately 0 . 3% and 1% for low and high titer infections , respectively . After blood feeding , unfed mosquitoes were removed . P . falciparum NF-54 gametocyte cultures were washed and mosquitoes were fed infected blood warmed to 37°C through a membrane feeder [42] . Post feeding , unfed mosquitoes were removed and blood-fed An . gambiae were incubated at 24°C for 7 days . Midguts of mosquitoes were dissected , stained with 0 . 2% mercurochrome and oocysts enumerated using a light contrast microscope ( Olympus ) . The Wolbachia density of each mosquito carcass was determined by qPCR as described above . The experiment was replicated 3 times . Replicate one was a high-gametocytemic culture , while replicates two and three had low gametocytemia . The variances of the data for replicates two and three did not differ statistically and were pooled for analysis ( squared ranks test , P>0 . 05 ) while replicate one was analyzed separately . Data were analyzed by Kruskal-Wallis test using the Dwass method for pairwise comparisons . An . gambiae female adults were injected with wMelPop or uninfected Mos55 cell culture homogenate ( control ) and fed a P . falciparum gametocyte infected or uninfected blood meal as previously described . Unfed mosquitoes were separated from fed mosquitoes . Mosquitoes were reared at 24°C at a density of approximately 30 mosquitoes per cup ( 4 cups per treatment ) and monitored twice daily for survival . Dead mosquitoes were removed from the experiment every 12 hours . The entire experiment was repeated twice . Data were analyzed by Kaplan-Meier analysis . Statistical significance was assessed by Kruskal-Wallis test using the Dwass method for pairwise comparisons .
|
Infection with Wolbachia bacteria has been shown to reduce pathogen levels in multiple mosquito species . Anopheles mosquitoes ( the obligate vectors of human malaria ) are naturally uninfected with Wolbachia , and stable artificial infections have not yet succeeded in this genus; however somatic infections can be established that can be used to assess the effect of Wolbachia infection in Anopheles . Here , we show that infection with two different Wolbachia strains ( wMelPop and wAlbB ) can significantly reduce levels of the human malaria parasite Plasmodium falciparum in Anopheles gambiae . After infection , Wolbachia disseminate throughout the mosquito but are notably absent from the gut and ovaries . The mosquito immune system is first induced in response to Wolbachia infection , but is then suppressed as the infection progresses . The Wolbachia strain wMelPop is highly virulent to Anopheles only after blood feeding . If stable infections can be established in Anopheles , and they act in a similar manner to somatic infections , Wolbachia could potentially be used as part of a strategy to control malaria .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"biotechnology",
"medicine",
"biochemistry",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"ecology",
"global",
"health",
"genetics",
"immunology",
"biology",
"microbiology",
"molecular",
"cell",
"biology",
"public",
"health",
"genetics",
"and",
"genomics"
] |
2011
|
Wolbachia Infections Are Virulent and Inhibit the Human Malaria Parasite Plasmodium Falciparum in Anopheles Gambiae
|
The role of secondary metabolites in the determination of cell identity has been an area of particular interest over recent years , and studies strongly indicate a connection between cell fate and the regulation of enzymes involved in secondary metabolism . In Arabidopsis thaliana , the maternally derived seed coat plays pivotal roles in both the protection of the developing embryo and the first steps of germination . In this regard , a characteristic feature of seed coat development is the accumulation of proanthocyanidins ( PAs - a class of phenylpropanoid metabolites ) in the innermost layer of the seed coat . Our genome-wide transcriptomic analysis suggests that the ovule identity factor SEEDSTICK ( STK ) is involved in the regulation of several metabolic processes , providing a strong basis for a connection between cell fate determination , development and metabolism . Using phenotypic , genetic , biochemical and transcriptomic approaches , we have focused specifically on the role of STK in PA biosynthesis . Our results indicate that STK exerts its effect by direct regulation of the gene encoding BANYULS/ANTHOCYANIDIN REDUCTASE ( BAN/ANR ) , which converts anthocyanidins into their corresponding 2 , 3-cis-flavan-3-ols . Our study also demonstrates that the levels of H3K9ac chromatin modification directly correlate with the active state of BAN in an STK-dependent way . This is consistent with the idea that MADS-domain proteins control the expression of their target genes through the modification of chromatin states . STK might thus recruit or regulate histone modifying factors to control their activity . In addition , we show that STK is able to regulate other BAN regulators . Our study demonstrates for the first time how a floral homeotic gene controls tissue identity through the regulation of a wide range of processes including the accumulation of secondary metabolites .
Seeds are essential units for plant propagation . Their development is an intricate process that requires the coordinated development of the embryo , the endosperm and the seed coat . The seed coat is derived from the maternal integuments and surrounds the embryo providing the latter protection against both mechanical damage as well as that inflicted by UV radiation . In addition , it facilitates the efficient dispersion of offspring and mediates initial water uptake during germination [1] . Upon double fertilization , the first phase of seed development is characterized by several morphological changes followed by the accumulation of secondary metabolites in specialized seed coat cells which mainly act in defence responses [2] . Although some interconnections between secondary metabolism and cell differentiation have been reported , the molecular mechanisms involved have not yet been elucidated . New approaches to genome-wide target identification indicate the existence of a correlation between cell identity specification and the regulation of secondary metabolism . For example , direct targets of the MADS-domain transcription factor SEPALLATA3 ( SEP3 ) have been shown to include genes involved in lipid biosynthesis , hormone production and the biosynthesis of sterol and wax [3]; SHORT VEGETATIVE PHASE ( SVP ) , another MADS-domain protein , binds to genes involved in the hormone stimulus response , suggesting its involvement in the cytokinin , auxin and jasmonate signalling pathways [4] . MADS-domain transcription factors have been demonstrated to be important regulators of floral organ specification , and SEEDSTICK ( STK ) in particular has been shown to play a pivotal role in ovule ontogeny [5] . The STK gene controls ovule identity redundantly with SHATTERPROOF1 ( SHP1 ) and SHP2 [6]–[8] . Furthermore , STK is required for normal seed shedding and , together with another MADS-domain gene ARABIDOPSIS B SISTER ( ABS ) , for proper formation of the endothelium , the innermost layer of the seed coat [7] , [9] . Transcriptome analysis of developing ovules and seeds has suggested the involvement of STK in the control of secondary metabolism ( see later ) . Based on the data obtained and the morphological characterization of the stk mutant we have focused on the regulation of the flavonoid metabolic pathway . Fertilized ovules accumulate proanthocyanidins ( PAs ) in the endothelium . These molecules are flavan-3-ols and in Arabidopsis they are composed of epicatechin monomers and polymers [10] . PAs are important compounds as they provide protection against light and predation by herbivores , they have antimicrobial and antioxidant activities , and in addition they limit the growth of neighbouring plants [11]–[16] . Epicatechins are responsible for the brown pigmentation of the Arabidopsis seed [16] . They are synthesized in the cytoplasm and then transported to the vacuoles where finally they are polymerized [17] . In Arabidopsis , PA biosynthesis begins in the micropylar region of the endothelium around 1 to 2 days after fertilization and then progressively extends to include the rest of the endothelium up to 5 to 6 days after fertilization [2] , [16] . The genetics of PA biosynthesis and accumulation has been well-studied in Arabidopsis revealing the participation of a complex network of several groups of genes . PA biosynthetic structural genes are divided into two groups: the Early ( EBGs ) and the Late ( LBGs ) Biosynthetic Genes [16] , [18] , [19] . The EBGs comprise four genes: CHALCONE SYNTHASE ( CHS ) , CHALCONE ISOMERASE ( CHI ) , FLAVONOL 3-HYDROXYLASE ( F3H ) and FLAVONOL 3′-HYDROXYLASE ( F3′H ) , and are involved in the biosynthesis of precursors for PAs and other classes of Arabidopsis flavonoids . The LBGs include DIHYDROFLAVONOL-4-REDUCTASE ( DFR ) , LEUCOANTHOCYANIDIN DIOXYGENASE ( LDOX ) and BANYULS/ANTHOCYANIDIN REDUCTASE ( BAN/ANR ) . BAN is considered to be a branch point in the phenylpropanoid biosynthetic pathway . It encodes an anthocyanidin reductase which converts anthocyanidins to their corresponding 2 , 3-cis-flavan-3-ols [20] . A third group of structural genes has also been identified in Arabidopsis that comprises TRANSPARENT TESTA 12 ( TT12 , MATE transporter ) , TT10 ( laccase 15 ) , TT19 ( glutathione-S-transferase ) and AHA10 ( H+-ATPase ) . These genes are involved in flavan-3-ol modification , transport and oxidation [21]–[24] and have recently been proposed to be LBG members [19] . The regulation of LBGs occurs via a ternary protein complex called MBW ( MYB-bHLH-WDR ) , formed by a specific R2R3-MYB , a bHLH transcription factor and the WD repeat protein TRANSPARENT TESTA GLABRA 1 ( TTG1; [16] , [25] , [26] ) . In addition to this complex , other transcription factors belonging to different families such as Zinc finger ( TT1/WIP1 ) , MADS ( ABS/TT16/AGL32 ) and WRKY ( TTG2/DSL1/WRKY44 ) also participate in their regulation [16] , [27]–[29] . Studies in several fields indicate a correlation between cell identity determination and metabolism . In unicellular organisms , like some species of yeast and Streptomyces , cellular differentiation is influenced by nutrients and metabolism [30] , [31] . Some years ago it was proposed that enzymes involved in carbon metabolism also regulate myoblast differentiation [32] , [33] , while stem cell differentiation has recently been proposed to be regulated by the metabolisms of both lipids and methionine [34] , [35] . The results presented in this manuscript demonstrate that STK directly prevents ectopic accumulation of PAs in the seed coat . STK both directly and indirectly regulates BAN and this involves STK-dependent changes in histone modification . The discovery of STK as a master regulator of genes involved in the anthocyanin biosynthetic pathway provides an interesting link between the determination of organ identity and more downstream cell-specific metabolic processes . Furthermore it opens up new possibilities to increase the levels of bioactive natural products which constitute a rich source of novel therapeutic compounds , or to modify pigments in plant tissues .
As a first step to understand which processes are controlled by STK , a high-throughput RNA-Seq analysis was performed comparing wild-type plants with the stk mutant . Based on the STK expression pattern , RNA was extracted from flowers starting from early stages of development ( stage 9 ) until maturity and after fertilization until 5 Days After Pollination ( DAP ) . Analysis of the raw data was performed on the commercially available CLC Genomics Workbench v . 4 . 7 . 1 ( http://www . clcbio . com/genomics/ ) . A total of 102 , 278 , 242 reads passed a quality filter and 85% were mapped back to the Arabidopsis TAIR10 genome . Approximately 90% of these mapped uniquely to single locations and each could thus be assigned to a single annotated TAIR10 gene . Normalization of expression was performed using RPKM values [36] . All other parameters were kept at default levels . The CLC Genomic Workbench was further used to identify and assess the levels of all the differentially expressed transcripts found in each cDNA library . Baggerley's test and False Discovery Rate ( FDR ) correction were used for the statistical evaluation of samples [37] . Our analysis revealed that 156 genes were up-regulated ( S1 Table ) in the stk mutant compared to wild type , whereas 90 were found to be down-regulated ( S2 Table ) . To obtain an initial insight into the potential functions of STK downstream genes , a global view of function and the underlying biology of the differentially expressed genes was obtained by examining their gene ontology using agriGO ( Fig . 1; [38] ) . For up-regulated genes in the stk mutant , the biological process category showed enrichment for lipid localization and secondary metabolic processes ( Fig . 1A; S3 Table ) . There was also notable enrichment for terms related to the phenylpropanoid metabolic process as well as flavonoid biosynthesis . Analysis of the molecular functions affected revealed enrichment for genes encoding pectinesterase , enzyme inhibitor and lipid binding activities ( Fig . 1B; S3 Table ) . Analysis of enriched cellular component terms in the list of up-regulated genes included categories related to the endomembrane system and also cell and cell part groups ( Fig . 1C; S3 Table ) . Among the group of genes down-regulated in stk , we only found significant enrichment of terms in the molecular function category related to DNA binding ( Fig . 1D; S3 Table ) . The transcriptome picture emerging provided a global view of the downstream networks regulated by STK . Interestingly , we found genes involved in flavonoid biosynthesis to be significantly represented in the up-regulated genes . That this group included key enzymatic players involved in PA synthesis constituted the basis for the work we report here on deciphering the role of STK in this seed coat process . The transcriptomic analysis of stk mutant ovules and seeds highlighted a significant increase in the abundance of transcripts involved in secondary metabolic processes , including those of genes involved in flavonoid and phenylpropanoid biosynthesis . Interestingly , in this group we found DFR , LDOX and BAN , three key enzyme encoding genes controlling the synthesis of catechin and epicatechin , the precursors of PAs . We therefore decided to study the role of STK in the regulation of the PA biosynthetic pathway in more detail since this is considered to be a key metabolic pathway linked to seed development ( for review see [16] ) . Furthermore , recent discoveries have shown that flavonoids play a fundamental role in regulating communication between the seed coat and the endosperm [39] . PAs in wild-type Arabidopsis seeds are accumulated in the endothelium . To investigate and compare the accumulation of PAs in wild-type and stk mutant seeds we made seed sections at the heart stage of embryo development ( when PA accumulation in the endothelium is completed ) and stained these with toluidine blue O . This staining provided a general view of all seed coat cells and revealed the presence of phenolic compounds in the endothelium of wild-type and stk mutant seeds as highlighted by the blue staining of their vacuoles ( Fig . 2A–C ) . However , the toluidine blue also evidenced considerable accumulation of phenolic compounds in the outermost layer of the inner integument ( ii2 ) in the stk mutant seed coat ( Fig . 2C , asterisk ) . In order to confirm the nature of the latter we used the vanillin assay that specifically detects flavan-3-ols and their proanthocyanidin polymers [2] . This analysis confirmed that PAs are accumulated in the endothelium in wild-type seeds ( Fig . 2D ) whereas in the stk mutant they are additionally observed in the ii2 layer ( Fig . 2E , asterisk ) . To better understand the role of STK in PA synthesis , soluble and insoluble extracts from mature and immature ( 6 DAP ) seeds were analyzed by Liquid Chromatography-Mass Spectrometry ( LC-MS ) and the complete metabolic profiles obtained are shown in Fig . 3 . Peaks that could be identified as known compounds were selected and analyzed ( S4 Table ) . No differences were observed in the metabolic profiles of insoluble PAs between mature wild-type and stk seeds ( S1 Figure ) . At 6 DAP , insoluble PAs could not be detected which concords with the solvent-soluble nature of PAs at the immature stage [10] . Soluble PAs , however , showed some differences: at 6 DAP , both wild-type and stk mutant seeds contained the same levels of PA oligomers ( n = 2–9 ) but the level of epicatechin monomers was higher in the stk mutant compared to the wild type ( Fig . 3A ) . In mature seeds only six soluble PA metabolites were detected . The levels of dimers , trimers and tetramers were the same in the wild type and in the mutant , but the levels of epicatechin monomers , pentamers and hexamers were different ( Fig . 3B ) . In particular , the levels of pentamers and hexamers in the wild type were higher compared to the mutant . By contrast the level of epicatechin monomers was higher in the stk mutant , as already detected in the metabolic profiles of immature seeds . The total amount of PAs in the stk mutant was greater than in wild-type seeds . These data support the morphological analysis and suggest that stk mutant seeds have a higher level of PAs than wild-type seeds . Furthermore , it implies that STK is predominantly involved in epicatechin monomer metabolism and only slightly affects oligomer production . The biosynthesis of PAs is dependent on structural genes that can be divided into EBGs and LBGs ( Fig . 4 ) , the EBGs being expressed prior to the LBGs [16] . Based on the transcriptome data , we focused our attention on the set of genes controlling the transformation of anthocyanidin into epicatechin ( Fig . 4 ) . The RNA-Seq data revealed that the expression of genes belonging to the EBGs is unaltered in the stk mutant background ( Table 1 ) ; however , the levels of all the LBGs were found to be increased in the mutant ( Table 1 ) . This suggests that STK acts as a repressor of the expression of all the LBGs . Among these , BAN codes for the core enzyme of PA production [25] , [40]–[42] . We performed quantitative Real Time-PCR ( qRT-PCR ) experiments on siliques from 0 to 6 DAP and confirmed the results obtained by the RNA-Seq experiment showing BAN to be up-regulated in the stk mutant ( S2 Figure ) . To investigate in which seed tissues BAN is expressed we performed in situ hybridization experiments using an antisense BAN probe . In wild-type seeds BAN was expressed only in the endothelium layer ( Fig . 5A; [20] ) . In contrast , BAN expression was observed in the endothelium layer and ectopically in the ii2 layer in the stk mutant ( Fig . 5B–C ) . These data confirm that STK controls the spatial and temporal expression of BAN . To investigate how STK regulates BAN expression we examined the expression pattern of the STK protein . We cloned the entire STK genomic region ( a DNA fragment comprising 3 . 5 kb of sequence upstream of the ATG codon plus all the coding region ) as a translational fusion to a GFP reporter gene , and the final pSTK::STK-GFP construct was introduced into the stk mutant background . The resulting plants produced seeds that were indistinguishable from wild-type and were able to abscise from the fruit upon maturity , demonstrating that the STK-GFP fusion protein was biologically active and able to fully complement the absence of the endogenous STK protein . STK-GFP was uniformly detected in the nuclei of the placenta and the early ovule primordia ( Fig . 6A–B ) which is consistent with previous data showing STK mRNA expression from stage 9 of flower development [7] . Later , as the developing ovules initiate the inner and outer integuments , GFP expression was found to be restricted to the nucellus and the funiculus ( Fig . 6B ) . However , during subsequent stages the fluorescence signal appeared throughout the outer and inner integuments ( Fig . 6C–D ) , and as development proceeded it could be seen to be present strongly in the growing funiculus and also in its contact region with the placenta . This pattern of protein localization during ovule development is consistent with previous in situ hybridization data [7] . During early embryogenesis ( Fig . 6E ) the STK-GFP protein remained extended throughout the outer integuments . Cell wall specific staining with propidium iodide ( PI ) was also carried out and allowed us to better define the sites of STK-GFP localization as the outer integuments and the second layer of the inner integument ( Fig . 6F ) . Our observations thus show that the distribution of the STK protein changes over the course of ovule and seed development . During early stages it is detected uniformly distributed in the placenta and the different cell types of the ovule primordia , whereas later it becomes restricted to the maternal seed coat and funiculus . No STK-GFP was detected in the ii1 and ii1′ layers suggesting that here STK does not repress BAN in these tissues . These data are consistent with the observation that STK is able to repress BAN expression only in the ii2 layer of the inner integument . RNA-Seq , expression analysis and in situ data all indicate that BAN expression is regulated by STK . In order to investigate whether this regulation involved direct interaction between the STK protein and the BAN gene we performed a ChIP ( Chromatin Immunoprecipitation ) assay on wild-type inflorescences and siliques ( up to 6 DAP ) using an antibody specific against the STK protein . Chromatin extracted from wild-type leaves was used as a negative control since STK is not expressed in this tissue , and the binding of STK to the VERDANDI ( VDD ) promoter [43] was used as the positive control . Bearing in mind that MADS-domain proteins recognize and bind CArG boxes [44] , the BAN gene genomic sequences comprising the 3 Kb upstream of the ATG start codon , the structural gene and 1 kb downstream of the STOP codon were analysed for the presence of such consensus motifs ( allowing up to one base mismatch; Fig . 5D ) . Six CArG boxes were found in the region of the BAN promoter sequence [2] extending up to 355 bp upstream from the ATG . We detected significant enrichment for the region immediately upstream of the BAN translational start site ( primer set spanning positions −140 to +1 , including three CArG boxes ( Fig . 5E ) ) thus indicating the presence of STK bound to this region . To investigate a possible mechanism of regulation of the expression of BAN by STK , we examined epigenetic marks at the BAN locus . Modifications such as the hyperacetylation of histones H3 and H4 have diverse impacts on gene transcriptional activity and chromatin organization [45] . H3K9ac is one of the most well-characterized epigenetic marks associated with active transcription and has been shown to influence numerous developmental and biological processes in higher plants [46]–[48] . To address the question of whether the differential expression of BAN observed between stk and wild-type tissues correlates with alterations in this epigenetic marker we analyzed wild-type and stk mutant siliques at 3–4 DAP for H3K9 acetylation at the BAN locus . ChIP experiments were performed using an antibody specific to H3K9ac and were analyzed by qRT-PCR ( Fig . 5F ) . IAA8 was used as a reference as it carries the H3K9ac mark and is equally expressed in both wild-type and stk mutant plants [49] . Interestingly , in wild-type material we found considerable enrichment of DNA sequences corresponding to the region around the translational start site within which STK binding sites are located . When we assayed the same region in the stk mutant we observed a dramatic increase in enrichment compared to wild type . These results are consistent with the presence of elevated levels of H3K9ac ( compared to H3 ) in the region of the wild-type BAN promoter where STK interacts ( transcriptionally active chromatin ) , and in addition evidences a very considerable enrichment of H3K9ac in this region in the stk mutant which correlates with the increased transcriptional activity of the BAN gene observed in the mutant background . Our data suggest that STK might regulate BAN expression directly binding to its promoter . Previous studies have identified other key regulators of BAN expression , including ABS , TT8 and ENHANCER OF GLABRA3 ( EGL3 ) . Whereas TT8 and EGL3 act redundantly in a protein complex that promotes the expression of BAN , ABS is necessary for PA biosynthesis and normal endothelium cell morphology [9] , [28] , [41] , [42] , [50] . This raised the question as to whether STK might act as master regulator also controlling the genes encoding transcription factors that regulate BAN . We therefore investigated the expression of ABS , EGL3 and TT8 in both wild-type and stk mutant siliques ( 3–4 DAP ) by qRT-PCR , and also analyzed the expression of ABS in developing flowers in order to study the relationship between STK and ABS in their roles in endothelium formation [9] . This experiment revealed up-regulation of TT8 , EGL3 and ABS in stk mutant siliques and hence that STK represses these genes in the wild type at this specific stage of development ( Fig . 7A ) ; that the level of ABS in un-pollinated inflorescences was unaffected indicates that STK does not play a role in regulating the expression of ABS prior to pollination . These data therefore suggest that STK regulates these genes during a narrow stage of development , at 3–4 DAP . In order to clarify whether STK directly controls the expression of these regulatory genes , we performed a ChIP assay . Bioinformatics analysis of the gene loci revealed the presence of two CArG boxes in the putative EGL3 promoter region at −2213 and −2177 bp; three CArG boxes in the putative ABS promoter region between −1692 and −1599 bp , and the TT8 gene presented two CArG boxes in the last exon at +3805 and +3815 bp from the ATG ( Fig . 7B ) . The ChIP assays showed that STK indeed binds to the CArG-containing regions of the EGL3 and ABS genes but not to the selected region of TT8 ( Fig . 7C ) . This analysis suggests that the control of PA biosynthesis by STK might occur via two mechanisms: a direct interaction with the BAN locus , and in addition by direct and indirect regulation of the expression of genes that encode transcriptional regulators of BAN .
Seed development is a highly complex process that includes the formation of the zygote , storage tissues and a protective seed coat . Differentiation of the various structures is evidenced at the morphological level but is also reflected by the spatial distribution of metabolites . STK is involved in seed development in several ways , for example stk mutant seeds are smaller than those of wild type and do not detach from the mother plant . Furthermore , combining the stk mutant with the arabidopsis b-sister ( abs ) mutant resulted in ovules that failed to develop the endothelium layer that forms the innermost component of the seed coat [9] . To gain a deeper insight into the role of STK during ovule and seed development we performed RNA-Seq on material extracted from wild-type and stk mutant inflorescences and seeds . The resulting transcriptomic analysis yielded a list of 246 genes identified as being differentially expressed between stk and wild type . The majority of these ( 156 genes ) were up-regulated in the stk background whilst 90 were down-regulated , which suggests that STK may act primarily as a repressor . Of all the deregulated transcripts , we found 18 that were expressed in the stk mutant but not in wild type , and 11 that were specifically expressed in the wild type but not in stk . GO analysis of the genes that were down-regulated in the stk mutant background revealed that they encoded proteins involved in DNA binding , including group II WRKY transcription factors ( TFs ) such as WRKY39 [51] . The importance of TTG2 ( WRKY group I ) as a regulator of the LBG biosynthetic pathway has been described previously [16] . Our RNA-Seq data revealed that expression levels of TTG2 are unaltered between wild type and stk . WRKY TFs are global regulators acting at various levels , including the direct modulation of immediately downstream target genes , but they also appear to interact with key chromatin-remodelling factors [52] . The latter could be of special interest since we found that the STK-dependent BAN regulatory mechanism involves a chromatin remodelling activity which may imply the action of downstream STK targets acting as remodelers that have yet to be defined . In this regard , the nucleosome assembly/disassembly protein NAP1-RELATED PROTEIN 1 ( NRP1 ) which interacts with chromatin remodelling factors [53] was also found in this list . In addition we identified NAC transcription factor-like 9 ( NTL9 ) , a calmodulin-regulated NAC transcriptional repressor in Arabidopsis [54] , and SQUAMOSA PROMOTER BINDING PROTEIN-LIKE 2 ( SPL2 ) , a member of the SQUAMOSA PROMOTER BINDING PROTEIN ( SBP ) -box family of transcription factors [55] . Future analysis will be directed to determine the possible mechanism ( s ) of action of STK . GO analysis of genes that are up-regulated in the stk mutant showed an abundance of transcripts involved in secondary metabolic processes , including those involved in flavonoid and phenylpropanoid biosynthesis . We hypothesize that STK may act on the PA pathway at two different levels: i ) impacting directly on the elements comprising the enzymatic pathway of PA biosynthesis , and/or ii ) influencing key regulatory elements . In the first case , all the genes of the Late Biosynthetic step were found to be up-regulated in the stk mutant . However , no changes were found for the group of genes involved in the metabolic steps corresponding to early synthesis . This may indicate that STK regulation does not involve major alterations at the level of production of dihydroflavonols , the precursors to PAs , but impact instead on the synthesis of anthocyandins and proanthocyanidins . We also found that two out of the three genes involved in the pathway controlling transport and compartmentation were up-regulated: TT12 and TT19 . Since anthocyanins are transported from the cytosol to the vacuole as part of the PA biosynthetic pathway , it may be reasonable to expect coordinated regulation of transporter-mediated and vesicle-mediated mechanisms by STK . As commented above , STK may also act on key regulatory elements . In this regard we observed a strong influence on the expression levels of the members of the bHLH transcription factor family ( TT8 and EGL3 ) , the Zinc-Finger transcription factor ( TT1 ) and the R2R3 MYB domain putative transcription factor ( TT2 ) in the stk mutant background . These regulatory elements together with other key MADS-domain transcription factors like ABS were all found to be up-regulated . Based on observation from the RNA-Seq data we investigated the role of STK in PA biosynthesis in more detail . Histological analysis revealed that PAs accumulated ectopically in the outer-layer of the inner integument in stk mutant seeds , and analysis by LC-MS analysis demonstrated that the level of soluble epicatechin monomers was greater in the stk mutant in both immature and mature seeds . The latter data confirm the morphological analysis ( Fig . 2 ) and support the idea that STK is involved in PA accumulation and in particular the accumulation of epicatechin monomer . Regulation of the production of flavonoids and proanthocyanidins has been extensively studied ( for review see [16] ) . Several transcription factors are known to be involved in the regulation of BAN expression , and in particular a R2R3-MYB/bHLH/WDR complex is responsible for BAN activation in the endothelium [25] , [40]–[42] . Not all the members of this complex exhibit the same hierarchy , for instance it has been suggested that TT8 is necessary for activation of the LBGs whereas a predominant role of TT2 controlling via a positive feedback loop is required to maintain the transcript levels of both TT8 and BAN [2] , [25] , [56] . Interestingly our RNA-Seq dataset highlighted a repressive role for STK on the regulation of TT2 . Another gene required for the activation of BAN in the endothelium is ABS . This gene is a member of the MADS-domain gene family and to date it is the only MADS-domain gene identified to be involved in the control of PA accumulation . abs mutants show significant reductions in epicatechin and procyanidins accumulation [10]; moreover , in the abs background both GUS reporter gene expression driven off the BAN promoter and qPCR analysis of BAN expression itself showed significant down-regulation [28] , [57] . In the complex regulatory network that governs PA accumulation in the endothelium layer , STK plays a key role controlling BAN expression , and our results suggest that this occurs through the binding of STK to the BAN regulatory region . Recently , Dean and collaborators performed genome-wide expression profiling using microarrays to identify those genes differentially expressed in the wild-type and abs mutant seed coats [57] . This showed that STK expression is unaffected in the abs mutant at 3 DAP but is up-regulated at 7 DAP . Furthermore , Nesi and colleagues demonstrated that ectopic expression of ABS causes altered PA accumulation in a manner very similar to that which we have observed in the stk mutant , since these plants present PAs not only in the endothelium but also in the more external layer of the inner integuments [28] . Integration of previous data with our results indicates that STK is a master regulator of inner seed coat differentiation . Whilst several transcription factors involved in seed coat determination or secondary metabolite accumulation have been already characterized [58] , no gene connecting these processes has been described previously . In our study we provide evidence of a role for chromatin modification , specifically H3K9 acetylation , in the transcriptional regulation of BAN . We have shown that the region of the BAN promoter proximal to the translational start site is heavily covered by this epigenetic marker of transcriptional activity . H3K9ac enrichment is greater when BAN is ectopically expressed due to the lack of STK protein . This suggests that STK somehow represses the activity of histone deacetylases ( HDACs ) at the BAN locus in cells of the outer layer of the inner integument . It will be interesting to determine whether STK is able to recruit a chromatin remodelling partner , yet to be identified , that would form part of a hypothetical STK complex . It is known that histone acetyltransferases ( HATs ) and HDACs participate in the genome-wide turnover of acetyl groups , and that besides histones some also modify other factors . Future progress will therefore be focused on determining their availability for interaction with specific transcription factors like STK and other protein complex partners . PA biosynthesis and its spatial accumulation are under the control of a complex regulatory network . BAN is one of the key genes of the LBG group and its expression falls under the influence of several transcription factors , including ABS , EGL3 and TT8 . Our work adds STK to this list . Our data also suggest STK to be a master regulator of PA biosynthesis and accumulation since we observed that it also controls the expression of ABS and EGL3 . This shows that STK not only acts as a regulator of ovule identity but also orchestrates important aspects of seed development , adding new evidence for the importance of MADS-domain genes in the control of plant developmental processes . The role of STK in the regulation of epicatechin accumulation could have relevance for certain agricultural applications: PAs are important in several aspects of plant protection and their significance in the flavour and astringency of foods and beverages is already known ( for review see [59] ) . Indeed , it has been demonstrated that avocado fruits that contain higher levels of epicatechin exhibit stronger resistance to fungal attack [60] . In this regard it will be interesting to study the fungal resistance of stk mutant seeds since regulation of STK levels might provide a tool to make plant seeds more resistant to fungi .
Arabidopsis thaliana wild-type ( ecotype Columbia ) and stk mutant plants were grown at 22°C under short-day ( 8 h light/16 h dark ) or long-day ( 16 h light/8 h dark ) conditions . The Arabidopsis stk mutant was kindly provided by M . Yanofsky [7] . The stk-2 allele contains a 74 nucleotide insertion near the splice site of the third intron . Identification of STK wild-type and mutant alleles was performed by PCR analysis using oligonucleotides AtP_204 ( 5′-GCTTGTTCTGATAGCACCAACACTAGCA-3′ ) and AtP_561 ( 5′-GGAACTCAAAGAGTCTCCCATCAG-3′ ) . The mutant allele yields a 399 bp DNA fragment whilst the wild-type allele produces a 325 bp fragment . Total RNA was extracted from two biological replicates ( 1 gr ) from both wild-type and stk mutant inflorescences and siliques until 5 DAP using the Qiagen ‘RNeasy miniKit’ according to the manufacturer's instructions . DNA contamination was removed using PROMEGA RQ1 RNase-Free DNase according to the manufacturer's instructions . RNA quality and integrity were analyzed by gel electrophoresis and validated on a Bioanalyzer 2100 ( Aligent , Santa Clara , CA ) ; RNA Integrity Number ( RIN ) values were greater than 7 for all samples . In order to confirm that the stk mutant was a knock-out line , STK expression was checked by qRT-PCR with primers RT_780 ( 5′-TGCGATGCAGAAGTTGCGCTC-3′ ) and RT_781 ( 5′-AGTACGCGGCATTGATTTCTTG-3′ ) . Sequencing libraries were prepared according to the manufacturer's instructions using the TruSeq RNA Sample Prep kit ( Illumina Inc . ) and sequenced on an Illumina HiSeq2000 ( 50 bp single-read ) . The processing of fluorescent images into sequences , base-calling and quality value calculations were performed using the Illumina data processing pipeline ( version 1 . 8 ) . Raw reads were filtered to obtain high-quality reads by removing low-quality reads containing more than 30% bases with Q<20 . Finally , quality control of the raw sequence data was performed using FastQC ( http://www . bioinformatics . babraham . ac . uk/projects/fastqc/ ) . Evaluation and treatment of raw data was performed on the commercially available CLC Genomics Workbench v . 4 . 7 . 1 ( http://www . clcbio . com/genomics/ ) . After trimming , the resulting high-quality reads were mapped onto the Arabidopsis genome ( TAIR10 ) . Approximately 25M reads of each sample that mapped with ≤2 mismatches were used for further analyses . The read number of each gene model was computed based on the coordinates of the mapped reads . A read was counted if any portion of that read's coordinates were included within a gene model . As CLC Genomics Workbench v . 4 . 7 . 1 distributes multireads at similar loci in proportion to the number of unique reads recorded and normalized by transcript length , we included both unique reads and reads that occur up to 10 times in the analysis to avoid undercount for genes that have closely related paralogs [36] . Gene expression values were based on reads per kilobase of exon model per million mapped read ( RPKM ) values [36] . The fold change and differential expression values between wild type and the stk mutant was calculated in terms of RPKM of the corresponding transcripts . Statistical analysis of biological replicates was assessed using heat map visualization of Euclidean distances . This clustered the biological replicates in the wild-type and mutant groups as expected . We further checked whether the overall variability of the samples reflected their grouping by performing Principal Component Analysis ( PCA ) . This confirmed that the replicates were relatively homogenous and distinguishable from the samples of the other group . Finally , the overall distribution of expression values between the different samples confirmed that none of the samples stood out from the rest . To obtain statistical confirmation of the differences in gene expression , P and FDR values were computed using Baggerley's test on expression proportions . We applied a threshold value of P = 0 , 05 to ensure that differential gene expression was maintained at a significant level ( 5% ) for the individual statistical tests . Transcripts that exhibited an estimated absolute Fold Change ≥2 ( i . e . 2 mapped reads per kilobase of mRNA ) were determined to be significantly differentially expressed . To gain insight into the biological processes associated with the regulated genes , we determined which GO annotation terms were over-represented , in both the up-regulated ( S1 Table ) and down-regulated ( S2 Table ) lists . Gene set enrichment analysis was performed with the agriGO database [38] using the Singular Enrichment Analysis ( SEA ) . All RNA-seq files are available from the NCBI GEO database ( accession number GSE59637 ) . For the morphological analysis of integuments , Arabidopsis thaliana wild-type ( ecotype Columbia ) and stk mutant plants were fixed for Technovit 7100 embedding ( Heraeus Kulzer ) following the manufacturer's instructions . Sections of plant tissue ( 0 . 8 µm ) were stained in 0 . 5% ( w/v ) toluidine blue O . Samples were observed using a Zeiss Axiophot D1 microscope ( http://zeiss . com/ ) equipped with differential interface contrast ( DIC ) optics . Images were recorded with an Axiocam MRc5 camera ( Zeiss ) using the Axiovision program ( version 4 . 1 ) . The whole-mount vanillin assay for PA detection was performed as described previously [61] . Vanillin ( vanilaldehyde ) condenses specifically with PAs and flavan-3-ol precursors to yield a bright-red product under acidic conditions . Microscopic observations were performed as detailed above . Three biological replicates of immature and mature seeds ( 30 mg ) were frozen in liquid nitrogen and ground to a fine powder using an analytical mill ( IKA; A11 basic ) . The soluble PA fraction was extracted with 75% methanol∶water ( v/v ) containing 0 . 1% formic acid . The mixture was vortexed , sonicated for 30 min at room temperature , vortexed again and then centrifuged ( 13000 rpm , 10 min ) . Soluble PAs ( contained in the supernatant ) were analyzed by LC-MS . The pellet was washed with 50% methanol∶water , centrifuged ( 13000 rpm , 10 min ) , washed again with 100% methanol and centrifuged ( 13000 rpm , 10 min ) . Samples were dried in a speed vacuum for 1 hour at 30°C and then hydrolyzed with NaOH ( 2N ) for 15 min at 60°C . The mixture was vortexed and HCl ( 4N ) was added . To remove lipids hexane was added to the mixture and after centrifugation ( 13000 rpm , 10 min ) the upper phase was removed . The insoluble PAs were extracted three times with ethyl acetate , and the three extractions were combined and dried in a speed vacuum for 1 hour at 30°C . The pellet was dissolved in acetone∶water∶acetic acid ( 70∶29 . 5∶0 . 5 v/v/v ) , vortexed , sonicated for 15 min and centrifuged ( 13000 rpm , 10 min ) . This PA fraction was analyzed by LC-MS . The profiling of PAs in extracts of seeds was performed by MS analysis using the UPLC-qTOF instrument ( Waters High Definition MS System; Synapt ) with the UPLC column connected online to a photodiode array detector ( Waters , Acquity ) and then to the MS detector equipped with an electrospray probe . The separation of metabolites and detection of the eluted compound masses was performed as previously described [62]–[66] . For the construction of pSTK::STK-GFP , DNA fragments containing the 3 , 5 kb promoter region and the complete STK genomic region without the stop codon were amplified from wild-type genomic DNA and cloned into a pGreen II binary vector containing a GFP reporter gene cassette . For amplification we used the following primers containing attB1 and attB2 recombination sequences: AtP_3066 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTCCAACCAATATCACACCCTAAATAC-3′ and AtP_3067 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTCGTCCGAGATGAAGAATTTTCTTGTC-3′ . The resulting binary vectors were transformed into Agrobacterium tumefaciens by electroporation prior to stable transformation of plants using the floral dip method . Transformant lines were obtained using BASTA as a selection agent . Resistant transgenic plants showing strong GFP fluorescence were genotyped and those that complemented the stk phenotype were selected . Protein expression patterns were analyzed by Confocal Laser Scanning Microscopy ( CLSM ) . Fresh material was collected , mounted in 10 mg/ml of propidium iodide ( Sigma P-4170 ) in water and immediately analyzed . CLSM analysis was performed using a Leica TCS SPE with a 488 nm argon laser line for excitation of GFP fluorescence; emissions were detected between 505 and 580 nm . For PI fluorescence a 543 nm laser line was used and emissions were detected between 600 and 640 nm . Confocal scans were performed with the pinhole at 1 airy unit . Images were collected in the multi-channel mode and the overlay images were generated using the Leica analysis software LAS AF 2 . 2 . 0 . DIG-labelled RNA probes for detection and hybridization of BAN were prepared as previously described [67] . Sections of plant tissue were hybridized with digoxigenin-labelled BAN antisense probe amplified using primers AtP_4331 ( 5′-CGAGTAGC TTATCTCTCTCG -3′ ) and AtP_4332 ( 5′- TCAATCCTTTTGACTCGAAG -3′ ) . The genomic regions located 3 kb upstream of the ATG , 1 kb downstream of the stop codon , and in the exons and introns of the selected genes were analyzed to identify CArG box sequences with up to one base mismatch . ChIP experiments were performed in a modified version of a previously reported protocol [68] . The qRT-PCR assay was conducted in triplicate on three different biological replicates , with three technical replicates for each sample , and was performed in a Bio-Rad iCycler iQ optical system ( software version 3 . 0a ) . ChIP efficiency was determined using the third CArG box of the VDD gene as a positive control [43] . Fold enrichment was calculated using the formulae of a previously reported protocol [43] . For ChIP-based analysis of histone modifications , the following antibodies were used for immunoprecipitation: rabbit anti-histone H3 ( Sigma-Aldrich H0164 ) and rabbit anti-H3 acetyl K9 ( Upstate 07-352 ) and were handled in parallel to samples lacking antibody . qRT-PCRs were performed on input and immunoprecipitated samples and % of input was calculated . The signal obtained after precipitation with anti-H3K9ac antibody ( as indicated in the figure ) was normalized to the signal obtained by precipitation with an antibody to an invariant domain of histone H3 . For H3K9ac analyses , IAA8 ( At2g22670 ) was used as a reference as it carries the H3K9ac mark and is equally expressed in both samples [49] . Relative enrichment of Mu-like transposons was included as negative control . Sequences of oligonucleotides used for ChIP analyses are listed in S5 Table . qRT-PCR experiments were performed on cDNA obtained from siliques from 3 to 4 DAP and unpollinated flowers . Total RNA was extracted using the LiCl method [69] . DNA contamination was removed using the Ambion TURBO DNA-free DNase kit according to the manufacturer's instructions . The treated RNA was reverse transcribed using the ImProm-II reverse transcription system ( Promega ) . cDNAs were used as templates in the qRT-PCR reactions containing the iQ SYBR Green Supermix ( Bio-Rad ) . The qRT-PCR assay was conducted in triplicate on three different biological replicates , with three technical replicates for each sample , and was performed in a Bio-Rad iCycler iQ Optical System ( software version 3 . 0a ) . Relative transcript enrichment of genes of interest was calculated normalizing the amount of mRNA against different endogenous control fragments ( UBQ , ACT , PPa2 and SAND [70] ) . The difference between the cycle threshold ( Ct ) of the gene and that of the reference gene ( ΔCt = CtGENE – CtREFERENCE ) was used to obtain the normalized expression of that gene , which corresponds to 2−ΔΔCt . The primers used for this analysis are listed in S5 Table . Annotated sequences used in this article can be found in the GenBank/EMBL data libraries under the following accession number: SEEDSTICK , At4g09960; BANYULS , At1g61720 .
|
Plant secondary metabolites accumulate in seeds to protect the developing embryo . Using an RNA sequencing approach in conjunction with enrichment analyses we identified the homeotic MADS-domain gene SEEDSTICK ( STK ) as a regulator of metabolic processes during seed development . We analyzed the role of STK as a key regulator of the production of proanthocyanidins , compounds which are important for the pigmentation of the seed . STK directly regulates a network of metabolic genes , and is also implicated in changes occurring in the chromatin landscape . Our work demonstrates that a key homeotic transcription factor not only determines the identity of ovules but also controls metabolic processes that occur subsequent to the initial identity determination process , thus suggesting a link between identity determination and cell-specific ( metabolic ) processes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"plant",
"science",
"cell",
"biology",
"plant",
"development",
"plant",
"growth",
"and",
"development",
"plant",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"molecular",
"cell",
"biology",
"molecular",
"biology"
] |
2014
|
SEEDSTICK is a Master Regulator of Development and Metabolism in the Arabidopsis Seed Coat
|
There is an intense debate concerning whether selection or demographics has been most important in shaping the sequence variation observed in modern human mitochondrial DNA ( mtDNA ) . Purifying selection is thought to be important in shaping mtDNA sequence evolution , but the strength of this selection has been debated , mainly due to the threshold effect of pathogenic mtDNA mutations and an observed excess of new mtDNA mutations in human population data . We experimentally addressed this issue by studying the maternal transmission of random mtDNA mutations in mtDNA mutator mice expressing a proofreading-deficient mitochondrial DNA polymerase . We report a rapid and strong elimination of nonsynonymous changes in protein-coding genes; the hallmark of purifying selection . There are striking similarities between the mutational patterns in our experimental mouse system and human mtDNA polymorphisms . These data show strong purifying selection against mutations within mtDNA protein-coding genes . To our knowledge , our study presents the first direct experimental observations of the fate of random mtDNA mutations in the mammalian germ line and demonstrates the importance of purifying selection in shaping mitochondrial sequence diversity .
Mammalian mitochondrial DNA ( mtDNA ) has a high mutation rate and is inherited in a non-Mendelian manner only from the mother [1 , 2] . Though there are reports of mitochondrial recombination in mammals , it is thought to be quite rare , and it is currently not known whether this phenomenon would be at a sufficient frequency to leave a signature in the population [3–6] . This asexual mode of transmission should leave the mitochondrial genome vulnerable to mutational meltdown by Muller's Ratchet , a process leading to deleterious mutation accumulation in asexual , nonrecombining lineages . The bottleneck phenomenon , which was first proposed after observation of rapid fixation of mitochondrial DNA variants in Holstein cows [7 , 8] , allows for rapid exposure of variant mtDNAs to selection at the level of the individual [9] , and may thereby , at the level of the population , protect against mutational meltdown . The over 100 , 000 mtDNA molecules in the mammalian oocyte do not undergo replication through the early stages of embryogenesis [2] . Therefore , these maternally derived mtDNA molecules are segregated through cell division events in the developing embryo to generate primordial germ cells with approximately 950–1 , 550 mtDNA copies [10] . Replication of mtDNA is reinitiated as the primordial germ cells migrate and differentiate to generate oocytes transmitting mtDNA to the next generation [2 , 11] . The mtDNA bottleneck appears to result from the replication of only a small subset of the mtDNA molecules as the primordial germ cells differentiate to generate oocytes [10] . It has long been thought that animal mtDNA is an essentially neutral marker of sequence evolution [1] , but evidence of the selective constraints on mtDNA is accumulating . Studies of animal mtDNA sequence variation within natural populations or in interspecies comparisons consistently show the signatures of negative selection ( e . g . , [12–14] ) . For humans , a considerable amount of mtDNA sequence is available from individuals as a result of studies into human evolution and human mtDNA diseases [15 , 16] . Analyses of the variation in human mtDNA sequences have led to a debate whether random genetic drift ( dependent on demographic history ) , positive selection , or purifying selection is important in the transmission and maintenance of this variation . [17 , 18] . Consensus is forming that selection is an important part of mtDNA sequence variation in human mtDNA , but the strength and nature of this selection are unresolved [18] . Population-level studies detect signatures of purifying selection in mtDNA sequence variation [19–26] , but recent accumulated variation within human populations implies neutrality or weak selection on these variants [25] . Findings from studies of mtDNA mutation inheritance in families with mtDNA-associated disease are compatible with the occurrence of only very weak or no selection on these mtDNA mutations [17 , 27] . Positive selection facilitated by climatic variation has recently been proposed for human mtDNA [28–30] and would profoundly affect the reliability of mitochondrial molecular clocks and drastically alter our understanding of human divergences . In an attempt to elucidate the mechanisms of mammalian mtDNA segregation in the germ line , several groups generated transmitochondrial mouse strains carrying two distinct mtDNA sequences ( a condition known as heteroplasmy ) . These transmitochondrial mice are generated by embryo–cytoplast fusions and exhibit germ line segregation patterns explainable by random drift [31–34] . However , tissue-specific segregation patterns within the offspring imply strong nuclear–mitochondrial interactions and suggest that molecular mechanisms exist that could allow for strong selection of mitochondrial variants within offspring [33 , 35–37] . Unfortunately , the technical complexities of the transmitochondrial technologies have much limited their use by research groups , and so far only a few sequence variants have been investigated . The mtDNA mutator mice are homozygous for a knock-in allele ( PolgAmut/PolgAmut ) expressing a proofreading-deficient catalytic subunit of mitochondrial DNA polymerase [38] . These mice have a substantial increase in the levels of mtDNA mutation in all investigated tissues . The somatic mutations generated are evenly distributed along an amplified fragment of the protein-coding mt-CYB gene of mtDNA , and all three codon positions are mutated at equal frequency , though transition mutations were more frequently observed than transversions [38] . In this study , we took advantage of this high mtDNA mutation rate to study the transmission of random mtDNA mutations in the mouse germ line . Female lineages were derived from mtDNA mutator mice by continuous backcrossing , allowing us to isolate , segregate , and characterize germ line mtDNA mutations .
We used eight mtDNA mutator founder females to establish independent maternal lines through 13 F1 females . The breeding scheme used ( Figure 1 ) takes advantage of the bottleneck phenomenon and allowed us to segregate the mtDNA mutations on a wild-type PolgA nuclear background from generation N2 and onwards . Sequencing was conducted from N2 onwards to sample only animals of wild-type PolgA nuclear background , and because levels of individual mutations in mtDNA mutator mice and N1 animals were too low to be detected by the sequencing methods employed . We sequenced the entire mtDNA of 190 animals from generations N2 to N6 and identified 1 , 069 unique mutations ( Dataset S1 ) . The typical animal carried approximately 30 mtDNA mutations ( mean = 29 . 8 mutations , standard deviation = 9 . 2 ) . In each line , a large proportion of the identified mtDNA mutations ( 38 . 48% ) were transmitted to the descendents of that particular N1 female , similar to the propagation of mtDNA haplogroups in human pedigrees . Other mutations were only observed in the siblings of a single litter ( 17 . 98% ) or in a single mouse ( 43 . 54% ) , but not in their offspring . Consistent with purifying selection acting on the mtDNA of these mouse lines , synonymous mutations were observed more frequently than nonsynonymous mutations . The ratio of nonsynonymous substitutions per site to synonymous substitutions per site for the protein-coding regions gave a value of 0 . 6035 , signifying purifying selection against amino acid changes in the protein-coding genes ( values less than 1 . 0 signify purifying selection ) . When mutations that occur only in an individual or a single litter were removed from the dataset , the ratio dropped to 0 . 4617 . The McDonald-Kreitman test of neutral evolution [39] and the accompanying Neutrality Index [40] were calculated for our mtDNA mutator lines , and gave values consistent with excess polymorphisms within our mtDNA mutator lines compared to either Mus musculus molossinus or the NZB mouse strain mtDNA sequences ( see Table S1 ) . We found a strong decrease in the number of mutations at the first and second codon positions of the protein-coding genes when compared to the third codon positions ( Figure 2A and Table S2A ) . This distribution of mutations is a hallmark of purifying selection , because changes in the first and second codon position usually result in an amino acid substitution , whereas many third codon position changes do not . This purifying selection is strong and rapid as the same codon distribution bias is evident in the N2 generation ( Figure S1 ) . The observed nucleotide mutational bias in protein-coding genes varied significantly from those observed for the other sites in the mtDNA molecule ( chi-square contingency table , p = 0 . 0023 ) thus showing differential selection pressures on the protein-coding genes versus other sites ( Table 1 ) . We observed the same selective signature against first and second codon positions when we compared 21 mouse-strain mtDNA sequences obtained from GenBank ( Figure 2B ) and human mtDNA sequence data obtained from the mtDB database [16] ( Figure 2C and Table S2B ) . There was a similar level of reduction of first codon position mutations in comparison with third codon position mutations in mtDNA mutator lines ( 2 . 0-fold reduction; Figure 2A ) and humans ( 2 . 6-fold reduction; Figure 2C ) . This striking similarity is surprising because the mtDNA mutator strains have undergone selection for at most six generations , whereas human sequence variation is the consequence of a much larger number of generations to act on these less deleterious substitutions . This illustrates the speed and strength of the selection on the mtDNA and its importance in sculpting modern mtDNA variation in natural populations . It also demonstrates this experimental model can be a powerful tool in investigation of mtDNA evolution . The smaller number of nonsynonymous changes in the mouse strains ( Figure 2B ) in comparison with mtDNA mutator lines ( Figure 2A ) can probably be explained by the limited sampling of only one individual from each of the 21 different mouse strains . In addition , it should be emphasized that the mtDNA mutator mice have been exposed to the effects of purifying selection for only a few generations . We further investigated the observed selection on protein-coding genes in the mtDNA mutator lines by separating the mutations at 4-fold degenerate sites ( third codon positions for amino acids L2 , V , A , T , P , S1 , R , and G ) from all other protein-coding mutations . The 4-fold degenerate sites can mutate to any nucleotide without changing the encoded amino acid and should therefore be subject to less selective constraint than other protein-coding sites . Expected values were calculated based on an assumption of an equal distribution of the observed mutations of these two classes , across the genes and corrected for their coding size . The ratios between observed and expected mutation frequencies at 4-fold degenerate sites were approximately equal in all of the protein-coding genes except for mt-ND2 and mt-ATP8 ( Figure 3A and Table S3A ) . In contrast , mutations at the non–4-fold degenerate sites deviated profoundly from the ratios predicted by equal distribution of mutations ( Figure 3A ) . Contingency table analysis was carried out to detect changes in the proportion of 4-fold degenerate site mutations to other sites within the protein-coding genes . The 13 protein-coding subunits were grouped by the oxidative phosphorylation ( OXPHOS ) enzyme complex to which they belong . Only complexes III ( containing mt-CYTB ) and complex IV ( containing mt-CO1–3 ) showed statistically significant changes in the ratio of 4-fold to non–4-fold sites ( see Table 2 ) . After correcting for multiple tests , only the complex IV data remained significant . Interestingly , the mt-ATP8 and mt-ATP6 subunits appear to allow for excess changes at all sites relative to the expected values , though the ratio of 4-fold to non–4-fold sites did not vary significantly ( Figure 3A ) . Analyses of human mtDNA sequences have shown a similar occurrence of excess sequence variation in the mt-ATP8 and mt-ATP6 genes , particularly evident for the mt-ATP6 gene [23–25 , 28] . These previous reports lead us to investigate available human sequence data , and we found a strong selection against non–4-fold degenerate changes in mt-CO1 versus the weaker selection in mt-ATP6 , mt-ATP8 , and mt-CYTB ( Figure 3B and Table S3B ) . Observed mutations for each protein gene for mtDNA mutator lines and human population showed the same variation from expected in 11 of 13 cases ( Figure 3C ) . Thus , sequence variation in protein-coding genes of this experimental mouse model demonstrates similar patterns to those seen in human populations . In contrast to the patterns found in the protein-coding genes , we found higher levels of mutations in tRNA and rRNA genes in the mtDNA mutator lines ( Figure 4A ) in comparison to the levels in mouse strains and humans ( Figure 4B and 4C ) . There are several observations from human mtDNA disease that imply that tRNA genes may be subject to a less rapid form of purifying selection than that observed for the protein-coding genes of our mtDNA mutator mouse lines: ( 1 ) Population-level sampling has revealed an increase in recent mtDNA sequence variation within tRNA and rRNA genes in humans [30 , 41]; ( 2 ) 58 . 2% of the known pathogenic human mtDNA mutations are located in the tRNA genes , although these genes only occupy 9 . 1% of the genome [42] , implying that these changes are , at low levels of heteroplasmy , more compatible with life than some protein-coding mutations; and ( 3 ) disease-causing tRNA gene mutations reach high heteroplasmic levels , or sometimes must be homoplasmic , before the onset of disease [43 , 44] . A less acute , but equally important , mechanism of purifying selection appears to be acting on tRNA genes . This may explain why the corresponding mutations are not removed as rapidly from the mtDNA mutator mouse lines . A very low mutation rate was observed for the control region of mtDNA mutator strains ( Figure 4A ) , despite the fact that the control region sequences are normally the most variable regions in mtDNAs . A reduction in the number of mutations within the control region was also reported in the somatic tissues of the mtDNA mutator mice [38] , though a mechanism to explain this observation is still elusive .
We present experimental evidence for strong purifying selection against nonsynonymous mutations in protein-coding genes during maternal transmission of mutated mtDNA in the mouse . The drastic reduction of mutations in the amino acid changing first and second codon positions of protein-coding genes are a direct result of purifying selection against deleterious mtDNA mutations at some stage within the reproductive cycle of these mice . This bias occurs rapidly and is evident as early as the N2 generation . These findings have profound implications for our understanding of how mutated mtDNA is transmitted between generations . It is important to recognize that this strong purifying selection against nonsynonymous changes that we observe is likely to be a universal phenomenon in mammals , but the rapid nature of this selective force would render these mutations difficult to detect in population studies . These findings have profound implications for our understanding of how mutated mtDNA is transmitted between generations . Within studies of human mtDNA evolution , the observation is that many substitutions are not ancient changes shared deep within human haplogroups , but rather are new variants clustered within the tips of phylogenetic networks and found only in a small number of individuals . This implies they are mildly deleterious variants not yet selected against [23 , 26 , 30] . Studies of disease-causing mtDNA mutations show they are often heteroplasmic , and can be present at high levels without consequence for the carrier . However , once the levels exceed a specific threshold , the respiratory chain function will be impaired , causing a clinical phenotype [11] . Based on these observations , when using the mtDNA mutator mouse to study germ line transmission of mtDNA mutations , one could expect to observe the inheritance of high numbers of mutations at all sites in the early generations , which would eventually be removed from the mouse lines once their phenotypic thresholds had been crossed . Whereas the inheritance of the tRNA , rRNA , and third codon position mutations appear to be following this expected pattern ( see Figures 2 and 4 ) , this is not the behaviour of mutations at the nonsynonymous first and second codon positions in our mouse lines ( Figure 2 ) . The strongest signature of purifying selection can be observed within mt-CO1 and mt-CO2 , consistent with the very high levels of sequence conservation in these genes . The strength and speed of this purifying selection could have other effects on the mutation patterns observed in our model . The consensus view is that bi-parental recombination of mammalian mtDNA is at most extremely rare [3–6] and therefore selection acting at any one site in the mtDNA will affect the entire mtDNA molecule . The observed strong and rapid selection of mtDNA mutations could therefore also reduce the number of neutral variants observed , due to their linkage to deleterious mutations . This means that 4-fold degenerate sites or even noncoding mutations might not be the reliable measure of the mitochondrial neutral mutation rate . Such an underestimation of the mtDNA mutation rate using phylogenetic or population methods relative to pedigree-based observation has been reported previously [45–47] . If this is the case , the models based on this assumption require recalibration . This point is also important in interpreting the excess change observed for the mt-CYB , mt-ATP6 , and mt-ATP8 genes in mtDNA mutator lines . Similar gene-specific increases of mutations have been reported in human mtDNA , especially in mt-ATP6 [23–25 , 28 , 29] . Though some argue that this signifies positive selection , the pattern may also be due to less-intense purifying selection on these specific genes . If mutations at mt-ATP6 experience less-selective constraint , mutations at these sites will be allowed to accumulate and persist in the mtDNA pool . Meanwhile , mutations at strongly selected sites , such as mt-CO1 and mt-CO2 , are eliminated , leading to the relative increase in the observed frequency of mt-ATP6 and mt-ATP8 mutations in our model organisms . In contrast to the rapid selection against nonsynonymous changes , rRNA and tRNA genes experienced less-intense purifying selection in our mtDNA mutator lines . Though tRNA genes also have high levels of sequence conservation , the frequency of observed mutations at these sites in our mouse lines was quite similar to the rate observed at third codon positions ( Figures 2A and 4A ) . Some of the identified tRNA mutations , e . g . , the deletion of one base in the anticodon loop of mt-TM ( 3873delC mutation ) can be predicted to have a biochemical effect if present at high levels . Previous models have mainly been based on observational studies of transmission of mutated mtDNA in human pedigrees affected by mitochondrial disease . Such threshold-mediated protection from selection should lead to slower purifying selection of the mtDNA variant , which may be reflected in the essentially neutral segregation patterns observed for disease-causing mutations prior to clinical manifestation [11 , 17 , 27 , 48] . It is plausible that these tRNAs , as well as a number of the nonsynonymous changes in the protein-coding genes in our model system , may eventually behave like human mtDNA disease mutations in that these mutations are transmitted and cause no obvious phenotype at low levels , but may be selected against or cause a disease-like phenotype at higher levels . We will continue sampling our lines to investigate the long-term fate of the observed tRNA gene mutations , as well as the stably transmitted nonsynonymous protein-coding gene changes . In the mtDNA mutator mice , the mutations within protein-coding genes are equally distributed across all three codon positions [38 , 49] , whereas the pattern of mutation accumulation is different in the mtDNA mutator lines . It has previously been proposed that mitochondrial fitness may be selected for during oocyte development [50] , and it is therefore quite possible that mtDNA in germ cells is under a different selective regime than the mtDNA in somatic cells . There is a massive proliferation of mtDNA during oogenesis , whereby a small number of mtDNA copies in the primordial germ cells are extensively amplified to generate the approximately 105 mtDNA copies in the mature oocyte [2 , 50] . This mechanism provides ample opportunities for functional testing of mtDNA during female germ-cell development , and future research is required to unravel molecular mechanisms responsible for this selection . Our experimental strategy has allowed us to look at the fate of a broad spectrum of mtDNA variation , and we report evidence of strong purifying selection in the mouse female germ line . All of the generated mtDNA mutator mouse lines showed the same strong reduction in nonsynonymous substitutions , exemplified by the reduction in first and second codon position mutations . This pattern is also seen in human populations and implies that purifying selection has a similar , drastic impact on the mtDNA variation in humans despite different demographic histories . The RNA genes , in contrast , appear to accumulate at levels approximating the synonymous third codon positions . These mutations are expected to eventually raise to high-enough levels and lead to impaired mitochondrial function in a manner similar to the threshold effect seen in human mtDNA disease . The data generated from this experimental model will allow us to build more accurate molecular models of mtDNA evolution and aid the understanding of inheritance patterns of human mtDNA disease mutations .
Heterozygous knock-in male mice ( PolgAmut/+ ) [38] were crossed to C57Bl/6NCrl females ( Charles River Laboratories ) . Resulting heterozygous mice were intercrossed to obtain female homozygous knock-in mice ( PolgAmut/PolgAmut ) . We performed crosses of mtDNA mutator females to wild type C57Bl/6 males to produce N1 females . Maternal mouse lines were then established by successive backcrossing of females to C57Bl/6 males ( Figure 1 ) . The genotype at the PolgA locus was determined as described previously [38] , and from the N2 generation onwards , only mice homozygous for the wild-type PolgA-allele were used in the study . In two cases , we bred heterozygous knock-in females from the N2 generation because of small litter sizes . In these two lines , all animals were homozygous for the wild-type PolgA-allele from generation N3 and onwards . This animal study was approved by the animal welfare ethics committee and performed in compliance with Swedish law . Pups were weaned at 21–25 d , and tissue from an ear punch was used for DNA isolation , as previously described [51] , except that the DNA was purified by phenol:chloroform:isoamyl alcohol ( 25:24:1 ) and chloroform extraction , followed by sodium acetate salt and ethanol precipitation . DNA was dissolved in an appropriate volume of deionised water . The mtDNA genome of each animal was amplified in 29 overlapping PCR reactions ( Table S4 ) . All PCR primers contain 5′ M13F or M13R sequence to use as sequencing primers . PCR samples were cleaned using the Agencourt AMPure PCR purification and directly sequenced in both directions using BigDye version 3 . 1 sequencing kit ( ABI ) . Cycle sequencing reactions were cleaned by precipitation ( ABI protocol ) or by the Agencourt CleanSEQ Dye-terminator removal kit . The sequence reactions were analysed on a 3130xl capillary sequencer ( ABI ) , and assembled and analysed for heteroplasmies and substitutions using SeqScape V2 . 5 software ( ABI ) and compared to our C57Bl/6 mtDNA reference sequence . The software identified heteroplasmy or substitutions at ≥25% signal intensity on both strands of DNA sequence . All heteroplasmies were confirmed by eye . All mutation sites are potentially heteroplasmic in this analysis , due to these detection thresholds in sequencing technology [52] . The wild-type C57Bl/6 mtDNA sequence used in this study varied from that presented in GenBank ( http://www . ncbi . nlm . nih . gov/Genbank ) accession number NC_005089 . 1 in two sites; position 9 , 461 was C ( synonymous change at amino acid 1 of mt-ND3 ) and position 11 , 515 was A ( S450N , nonsynonymous change in mt-ND4 ) . Two poly-A tracts in the mouse genome could not be reliably scored for insertion and deletions due to sequencing complications . We therefore ignored insertions/deletions at positions 5 , 160–5 , 191 ( origin of light strand replication ) and 9 , 821–9 , 830 ( polymorphic region of mt-tR ) . The following mouse sequences were used for comparison of strain variation , GenBank accession numbers: AB042432 , AB042523 , AB042524 , AB042809 , AB049357 , AJ489607 , AJ512208 , AY339599 , AY466499 , AY533105 , AY533106 , AY533107 , AY533108 , AY675564 , AY999076 , DQ106412 , DQ106413 , J01420 , L07095 , L07096 , and V00711 . Sequences were aligned to the wild-type C57Bl/6 mtDNA sequence , and all variations were scored . Human mtDNA sequence data were obtained from the mtDB human database ( http://www . genpat . uu . se/mtDB/; accessed February 2007 ) [16] . Variants were classified as all observed sequence changes from the most prominently observed nucleotide at the given position in the database . Codon usage for our C57Bl/6 line was calculated using CodonW version 1 . 3 ( John Peden , http://codonw . sourceforge . net// ) . The codon usage was used to calculate the number of synonymous and nonsynonymous substitution sites in the C57Bl/6 mtDNA genome . The number of 4-fold degenerate sites versus other sites for the protein-coding genes was derived from codon usage calculations on each protein-coding gene . Expect values ( Figure 3A and 3B ) made the assumption of a random distribution of observed mutations across the mtDNA molecule . The genome total of observed mutations within each class was multiplied by the proportion of those sites encoded for each gene . Observed mutations are reported as the number of mutations for that gene detected , multiplied by the proportion of sites for that class within that gene . Comparison of nucleotide biases was conducted using a contingency analysis on the data supplied in Table 1 . The analysis was performed on the eight substitution classes , where all observed values were greater than six . The variation was observed to be significant ( chi-square test , p = 0 . 0023 , 7 df ) . Comparisons of the 4-fold versus non–4-fold mutations for each mitochondrial OXPHOS enzyme complex were carried out using the Fisher exact test on a contingency table comparing each enzyme subunit to the sum of mutations for the remaining complexes . The Bonferroni correction for multiple testing of the data was applied when determining significance .
The GenBank ( http://www . ncbi . nlm . nih . gov/Genbank ) accession numbers for the mouse sequences used for comparison of strain variation to the wild-type C57Bl/6 mtDNA sequence ( DQ106412 ) are as follows: AB042432 , AB042523 , AB042524 , AB042809 , AB049357 , AJ489607 , AJ512208 , AY339599 , AY466499 , AY533105 , AY533106 , AY533107 , AY533108 , AY675564 , AY999076 , DQ106413 , J01420 , L07095 , L07096 , and V00711 .
|
Mammalian mitochondrial DNA ( mtDNA ) is maternally transmitted and does not undergo bi-parental recombination in the germ line . This asexual mode of transmission , together with a high rate of mutation , should eventually lead to the accumulation of numerous deleterious mtDNA mutations and a “mutational meltdown” ( a phenomenon know as Muller's Ratchet ) . In this study , we utilized a genetic mouse model , the mtDNA mutator mouse , to introduce random mtDNA mutations , and followed transmission of these mutations . Maternal transmission of mtDNA is typically subjected to a bottleneck phenomenon whereby only a fraction of the mtDNA copies in the germ-cell precursor are amplified to generate the approximately 105 mtDNA copies present in the mature oocyte . As a consequence of this phenomenon , the established maternal mouse lines carried high levels of a few mtDNA mutations . We sequenced the entire mtDNA to characterize the maternally transmitted mutations in the established mouse lines . Surprisingly , mutations causing amino acid changes were strongly underrepresented in comparison with “silent” changes in the protein-coding genes . These results show that mtDNA is subject to strong purifying selection in the maternal germ line . Such selection of functional mtDNA genomes likely involves a mechanism for functional testing to prevent transmission of mutated genomes to the offspring .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"evolutionary",
"biology",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] |
2008
|
Strong Purifying Selection in Transmission of Mammalian Mitochondrial DNA
|
Reversion and spread of vaccine-derived poliovirus ( VDPV ) to cause outbreaks of poliomyelitis is a rare outcome resulting from immunisation with the live-attenuated oral poliovirus vaccines ( OPVs ) . Global withdrawal of all three OPV serotypes is therefore a key objective of the polio endgame strategic plan , starting with serotype 2 ( OPV2 ) in April 2016 . Supplementary immunisation activities ( SIAs ) with trivalent OPV ( tOPV ) in advance of this date could mitigate the risks of OPV2 withdrawal by increasing serotype-2 immunity , but may also create new serotype-2 VDPV ( VDPV2 ) . Here , we examine the risk factors for VDPV2 emergence and implications for the strategy of tOPV SIAs prior to OPV2 withdrawal . We first developed mathematical models of VDPV2 emergence and spread . We found that in settings with low routine immunisation coverage , the implementation of a single SIA increases the risk of VDPV2 emergence . If routine coverage is 20% , at least 3 SIAs are needed to bring that risk close to zero , and if SIA coverage is low or there are persistently “missed” groups , the risk remains high despite the implementation of multiple SIAs . We then analysed data from Nigeria on the 29 VDPV2 emergences that occurred during 2004−2014 . Districts reporting the first case of poliomyelitis associated with a VDPV2 emergence were compared to districts with no VDPV2 emergence in the same 6-month period using conditional logistic regression . In agreement with the model results , the odds of VDPV2 emergence decreased with higher routine immunisation coverage ( odds ratio 0 . 67 for a 10% absolute increase in coverage [95% confidence interval 0 . 55−0 . 82] ) . We also found that the probability of a VDPV2 emergence resulting in poliomyelitis in >1 child was significantly higher in districts with low serotype-2 population immunity . Our results support a strategy of focused tOPV SIAs before OPV2 withdrawal in areas at risk of VDPV2 emergence and in sufficient number to raise population immunity above the threshold permitting VDPV2 circulation . A failure to implement this risk-based approach could mean these SIAs actually increase the risk of VDPV2 emergence and spread .
Global and synchronous withdrawal of all live-attenuated oral poliovirus vaccines ( OPV ) is one of the major objectives of the global Polio Eradication & Endgame Strategic Plan 2013–2018 [1] and part of the global transition from OPV to inactivated poliovirus vaccine ( IPV ) . Serotype 2 will be the first to be removed , with a planned date of April 2016 . This means that trivalent OPV ( tOPV ) will be replaced by bivalent OPV ( bOPV , containing Sabin virus types 1 and 3 ) in routine immunisation programmes , and tOPV will no longer be used in supplementary immunisation activities ( SIAs ) . Furthermore , all OPV-using countries are recommended to introduce at least one dose of IPV in their routine immunisation programmes before the switch from tOPV to bOPV [2] . OPV use needs to be stopped because of its genetic instability . Attenuated vaccine ( Sabin ) polioviruses lose key genetic determinants of attenuation through mutation and/or recombination with other enterovirus serotypes during replication in the human gut [3] . In countries using OPV , approximately 1 child per 900 , 000 first OPV doses is estimated to develop vaccine-associated paralytic poliomyelitis ( VAPP ) [4] . The relative contributions of viral evolution ( loss of key attenuating sites ) , immune function of the vaccine recipient and chance in the aetiology of VAPP are unclear . More significantly for the Global Polio Eradication Initiative ( GPEI ) , vaccine polioviruses may spread from the recipient to his or her contacts , in rare cases leading to an outbreak of a vaccine-derived poliovirus ( VDPV ) . VDPVs are defined as OPV-related isolates whose ~900-nucleotide sequence encoding the major capsid protein VP1 differs from that of the parental strain by >1% for serotypes 1 and 3 , and >0 . 6% for serotype 2 [5] . VDPVs are classified into three categories: circulating VDPVs ( cVDPVs ) , when there is evidence of person-to-person transmission; immunodeficiency-associated VDPVs ( iVDPVs ) , shed by individuals with primary immunodeficiencies who have prolonged , sometimes chronic , virus excretion; and , ambiguous VDPVs ( aVDPVs ) , which are isolates that cannot be classified as cVDPV or iVDPV despite thorough investigation [6 , 7] . Until July 2015 , the definition of cVDPV required that genetically linked VDPVs were isolated from at least two AFP cases , but the GPEI now considers even single individual or environmental sample isolates to be cVDPV if their genetic features indicate prolonged circulation [7] . cVDPVs have transmission dynamics similar to wild polioviruses [8] . Since 2006 , more than 680 acute flaccid paralysis ( AFP ) cases due to VDPVs have been reported worldwide [9] , underlining the importance of VDPVs for the polio eradication endgame . Strikingly , >97% of those cases have been associated with serotype 2 [9] , whose wild counterpart was last detected in 1999 [10] . The burden of serotype 2 VDPV ( VDPV2 ) and the eradication of serotype 2 wild poliovirus ( WPV2 ) in 1999 are the main motivations for the global withdrawal of serotype 2 OPV ( OPV2 ) planned for April 2016 . Polioviruses spread where levels of immunity in the population are low and where environmental conditions such as sanitation and crowding facilitate virus transmission . As such , the detection of cVDPVs has historically been associated with poor population immunity [3 , 8 , 11–15] . However , the initial appearance of a VDPV in a population depends on different factors and the relationship with population immunity may be more complex . For example , the number of people infected with Sabin poliovirus , the duration of excretion among those infected , the extent of secondary transmission and the prevalence of other enteroviruses may all be important in determining the probability of VDPV emergence . Worldwide OPV2 withdrawal will put the 155 countries currently using tOPV in their routine immunisation programmes at risk of outbreaks of VDPV2 given the associated increase in the number of children susceptible to that type . SIAs with tOPV prior to OPV2 withdrawal would increase population immunity to serotype 2 and have been proposed as a strategy to mitigate the risk of VDPV2 emergence and spread [16] . However , infrequent or poor-coverage SIAs could lead to limited immunity and potentially an adverse increase in risk resulting from poliovirus shedding and seeding of new VDPV . A better understanding of the factors associated with the risk of VDPV emergence and subsequent spread will help the GPEI to define a clear strategy on the number , timing and geographic extent of any tOPV SIA that minimises the risk of VDPV2 emergence at the time of and immediately after OPV2 withdrawal . Defining such a strategy is one of the priorities of the polio eradication program . In this article , we first present mathematical models that describe the relationship between the coverage of routine and supplementary immunisation activities , and the probability of VDPV emergence and subsequent spread . To test the conclusions from the mathematical models , we then identified risk factors associated with past VDPV2 emergences in Nigeria . For this aim , we carried out a case-control analysis of those districts reporting the first case of poliomyelitis associated with each of the 29 independent VDPV2 emergences in Nigeria during 2004−2014 compared with districts without emergences . We also used logistic regression to identify the risk factors associated with the probability that a VDPV2 emergence resulted in >1 case of poliomyelitis . We finish by discussing the implications of our findings for the tOPV SIA strategy to reduce the risk of VDPV2 emergence during and post OPV2 withdrawal .
The number of people infected with Sabin polioviruses is primarily determined by the number of doses of OPV administered during routine and supplementary immunisation activities and the level of population immunity . As the amount of OPV administered increases from zero , the number of Sabin-infected individuals will initially increase , but at some point further increases in OPV administration are likely to result in a decrease in the number of individuals infected because of the associated increase in the level of population immunity . This implies a trade-off in the levels of OPV use that will favour VDPV emergence . We developed two mathematical models to study this trade-off: an analytical model that includes only SIAs , and a more complex model that includes both routine immunisation and SIAs , which must be solved through numerical simulation . We used these models to investigate the risks and benefits of carrying out preventive campaigns with tOPV as a strategy to maximise population immunity to serotype 2 prior to OPV2 withdrawal .
We explored the probability of a VDPV outbreak for different numbers of supplementary campaigns in a scenario without routine immunisation and considering a population completely susceptible ( Fig 2 ) . Using the analytical model , if the individuals reached at each campaign are randomly chosen ( assuming the same coverage at each campaign ) , the risk of a VDPV outbreak is maximised at low and intermediate levels of SIA coverage ( Fig 2A ) . The exact location of the peak in risk depends on the number of SIAs , rapidly shifting to lower values of SIA coverage for increasing number of SIAs . In particular , for a single SIA with 100% coverage , the probability of a VDPV outbreak is around 70% , which is explained by the relatively small proportion of children that will be protected after the campaign , due to the limited ( ~50% ) immunogenicity of OPV . As expected , the size of any resulting outbreak is also significantly smaller for increasing number of SIAs ( Fig 2D ) . Although the absolute risk of VDPV emergence depends on the assumed probability of reversion of Sabin poliovirus to a VDPV ( ρ ) and the assumed population size ( N ) via σ = ρN , the location of the peak in risk does not change unless the value of σ is so low or high as to make VDPV emergence impossible or inevitable respectively . We provide a sensitivity analysis of the probability of VDPV outbreak to the value of σ in Fig B in S1 Text . In particular , when σ becomes sufficiently large , the probability of a VDPV outbreak becomes a stepwise function of SIA coverage ( Fig B in S1 Text ) . If the same individuals are reached at each campaign , thus leaving a “missed” group that is only immunised through secondary spread of Sabin virus , the risk of a VDPV outbreak is maximised at intermediate levels of SIA coverage . More importantly , increasing the number of SIAs above 4 barely reduces the risk , which becomes zero only above 70% coverage after 4 or more SIAs ( Fig 2B ) . In other words , there is a threshold in SIA coverage under which the risk does not decrease despite increasing the number of campaigns . The existence of this threshold can be shown analytically ( Section A . 2 . 3 in S1 Text ) , and for both random and fixed coverage , an expression for the minimum SIA vaccine coverage required to have zero probability of outbreak can be found ( Section A . 2 . 3 in S1 Text ) . A sensitivity analysis of the minimum SIA coverage required for zero probability of a VDPV outbreak to a broad range of values of the reproduction number of Sabin virus and the reproduction number of VDPVs is shown in Figs D and E in S1 Text . As expected , the minimum SIA coverage to bring the probability of a VDPV outbreak to zero increases for increasing values of the reproduction number of VDPVs , however , it slightly decreases for increasing values of the reproduction number of Sabin virus , because of the associated increase in the number of individuals who will be immunised by secondary spread of OPV from vaccinees . The risk of observing an outbreak of VDPV obtained with the more complex model in the absence of routine immunisation coverage displays a shape similar to that obtained with the analytical model , although stochastic extinction results in a lower risk at low values of SIA coverage ( Fig 2C ) . The stochastic SIR model allows the study of the risk of VDPV2 outbreak in the context of OPV2 withdrawal . Including reasonable levels of routine immunisation coverage results in a significant reduction in the risk of VDPV2 outbreaks during the 6 months that follow OPV2 withdrawal ( Fig 3A and 3B ) . This risk becomes almost negligible when routine coverage is high ( Fig 3B ) , but for low levels of routine immunisation coverage , multiple tOPV SIAs preceding OPV2 withdrawal are needed to avoid seeding new VDPV2 ( Fig 3C and 3D ) . Notably , in the context of low routine immunisation coverage , a single SIA seems to highly increase the risk irrespective of SIA coverage , and at least 3 campaigns at high coverage are needed to bring that risk close to zero ( Fig 3C and 3D ) . If campaign coverage is only intermediate and there is a persistently “missed” group ( i . e . SIAs reach the same individuals at each round ) , the risk of VDPV2 outbreak remains high even after 4 or 5 SIAs ( Fig 3D ) . A total of 29 independent VDPV2 emergence events were identified in Nigeria during the study period , of which 7 resulted in more than one case of poliomyelitis ( Fig 4A , Table C in S1 Text ) . This resulted in 28 cases in the case-control analysis , since two emergences took place in the same district during the same 6-month period ( Maiduguri , Borno state , between April and September 2006 ) . The 28 cases were matched to 560 controls . The number of tOPV SIAs in the previous 6 months varied between 0 and 5 ( Fig H in S1 Text ) . Important changes over time occurred , due to a progressive and rapid removal of tOPV from SIA since 2006 , which was replaced by bOPV , mOPV1 and mOPV3 [29] . tOPV was re-introduced in SIAs since mid-2009 . These changes over time were also reflected in estimated serotype-2 immunity among children 0–2 years old , which reached very low levels in 2008 and 2009 , and increased again from 2010 onwards ( Fig G in S1 Text ) . In general , serotype-2 population immunity was higher in Southern districts . Routine immunisation coverage was also higher in the South and increased during the study period ( Fig 4B ) . The annual number of births per district was highly variable , ranging from 42 children ( Bakassi , Cross River state ) to 57 , 710 children ( Alimosho , Lagos state ) , with a median of 7 , 542 ( Fig I in S1 Text ) . Population density was also highly variable , ranging between an average of 9 . 37 ( Teungo , Adamawa state ) and 55 , 450 people per km2 ( Ajeromi-Ifelodun , Lagos state ) , with a median of 218 . 70 ( Fig J in S1 Text ) . The mean number of household members per district remained nearly constant over the study period , ranging between 3 . 26 and 6 . 31 , and displayed a North-South gradient ( Fig Q in S1 Text ) . In the univariable analyses , a number of variables were associated with cases of VDPV2 emergence: ( i ) geographic region ( North vs . South ) , ( ii ) serotype-2 population immunity , ( iii ) routine immunisation coverage , ( iv ) number of tOPV campaigns in the previous 6 months , ( v ) number of months since the last tOPV campaign , ( vi ) number of births , and ( vii ) number of household members ( Table 1 ) . Districts in the North had an increased risk of VDPV2 emergence compared to the South ( odds ratio 5 . 52 , [95% confidence interval 1 . 89−16 . 16] ) . This association may well reflect the existence of a North-South gradient in Nigeria for many demographic , social and economic variables [30] . The number of births was also associated with cases of VDPV2 emergence as it is a proxy for the size of the population exposed to OPV . Interestingly , among the variables related to OPV use , increased population immunity , routine immunisation coverage and the number of months since the last tOPV SIA were associated with a reduced odds of VDPV2 emergence . However , the number of campaigns in the previous 6 months was associated with an increase in the odds of VDPV2 emergence ( Table 1 ) . The best multivariable model ( lowest AIC , 146 . 63 ) retained two variables statistically significantly associated with cases of VDPV2 emergence: routine immunisation coverage and the annual number of births ( Table 1 ) . In this model , an absolute increase of 10% in routine immunisation coverage was estimated to reduce the odds of VDPV2 emergence by 31% . Adding the number of tOPV SIAs in the previous 6 months to the best model gave a very similar AIC ( 148 . 08 ) , but the variable was not statistically significant ( odds ratio 1 . 54 , [95% confidence interval 0 . 47−5 . 07] ) . The seven VDPV2 emergences that established circulating lineages ( >1 case of poliomyelitis ) occurred in districts with low to middle serotype-2 population immunity ( <55% ) and low routine immunisation coverage ( <20% ) ( Fig 5 ) . A univariable logistic regression analysis found that the probability of an emergent VDPV2 to establish a circulating lineage decreased for higher serotype-2 population immunity ( p = 0 . 051 ) . The other variables did not show a statistically significant association with the probability of a VDPV2 to establish a circulating lineage .
This study presents an analysis of the risk factors associated with the emergence and spread of VDPV , and provides a basis for strategic decisions about the optimal extent and number of mass campaigns with OPV in advance of OPV withdrawal . First , using two simple mathematical models , we describe a trade-off between OPV use and the risk of VDPV emergence . Our findings indicate that immunity provided through routine immunisation counterbalances well the risk of VDPV emergence and spread . However , we found that in settings where routine immunisation coverage or the baseline level of population immunity is low , a small number of SIA campaigns could increase the risk of VDPV emergence compared to no campaigns . This is partly due to the low immunogenicity of OPV that makes necessary a certain number of SIA rounds to increase population immunity to levels that counterbalance the risk of seeding new VDPV through those campaigns . For example , our model predicted that in a setting with just 20% routine coverage with three tOPV doses ( e . g . many districts in northern Nigeria ) , a single OPV SIA increased the risk of VDPV outbreak irrespective of SIA coverage ( Fig 3 ) . To bring population immunity to levels that completely counterbalanced the risk of VDPV outbreak , at least three rounds of supplementary campaigns at 80% coverage were needed . If only intermediate levels of campaign coverage were attained and SIAs persistently reached the same population leaving a persistently “missed” group , the risk of VDPV emergence remained high even when a high number of campaigns were implemented ( Fig 3D ) . The existence of this threshold in SIA coverage under which the risk cannot decrease despite an increasing number of SIA illustrates how groups of unvaccinated children may hamper the efforts to minimise the risk of VDPV after OPV withdrawal . Second , we identified risk factors associated with VDPV2 emergence and subsequent spread in Nigeria using epidemiologic , virologic and demographic data for 2004−2014 . In both univariable and multivariable analyses , districts reporting the first case of poliomyelitis associated with a given VDPV2 emergence were more likely to have low routine immunisation coverage and a higher number of births . These districts were also more likely to have had a higher number of tOPV SIA in the previous 6 months , although this association was not statistically significant in the final multivariable model . This may be a result of the small number of observations , or a confounding between routine immunisation and tOPV campaigns , which are used to fill the immunity gaps . Finally , we also found that VDPV2 emergences were more likely to establish a circulating lineage and thus be responsible for more than one AFP case when they emerged in districts with low serotype-2 population immunity . Interestingly , the VDPV2 emergences that established a circulating lineage ( 7/29 ) occurred in districts with estimated serotype-2 population immunity <55% . The statistical analyses of data on VDPV2 emergence in Nigeria are consistent with our transmission model results in suggesting that tOPV SIAs can in some settings increase the risk of VDPV emergence . Past experience has also highlighted how limited use of OPV ( Sabin or other attenuated strains ) either during small clinical trials ( e . g . Poland [31 , 32] ) or vaccination programmes ( e . g . Byelorussia , former USSR [33] ) can lead to widespread circulation of VDPV and outbreaks of poliomyelitis [31–33] . In Nigeria , the association that we found could be explained either by an insufficient number of campaigns , low coverage of those campaigns ( settings with a higher number of campaigns may have poorer campaign coverage ) or the existence of persistently “missed” populations , leading to insufficient levels of population immunity to avoid seeding new VDPV . Introduced in 2009 to monitor the quality of SIAs , lot quality assurance sampling ( LQAS ) showed that SIA coverage in 65% of districts in Nigeria did not reach 60% by the end of 2009 [34] , suggesting that possibly only intermediate levels of SIA coverage were reached during the first half of the study period . Promisingly , significant improvements in SIA coverage have been reported since [34] . There are several limitations to our analyses . Firstly , the results of the case-control analyses were limited by the small number of emergence events in Nigeria , resulting in wide confidence intervals for some odds ratios . However , we chose Nigeria because it has experienced the greatest number of recorded VDPV2 emergences , with each VDPV2 undergoing detailed genetic sequencing and molecular epidemiological analysis [23] . Secondly , we were only able to examine risk factors associated with the first reported case of poliomyelitis caused by a VDPV2 , rather than the initial emergence of the VDPV2 . These VDPV2 isolates had between 6 and 17 nucleotide substitutions in the VP1 coding region , corresponding to an estimated average time of circulation since the initiating tOPV dose of around 9 months [23] , thus leaving the possibility that the district where the first AFP case associated to a given emergence was detected did not correspond to the district where the initiating tOPV dose was administered . Thirdly , although our simple mathematical model is mechanistic–describing OPV transmission and VDPV emergence–it does not attempt to capture the detailed genetic changes that result in reversion of Sabin poliovirus to a VDPV with transmissibility and virulence equivalent to wild-type virus [8] . Attenuating mutations in the Sabin polioviruses have been identified , but the process of reversion and the significance of genetic changes for poliovirus transmission are not well understood [3] . Instead , we chose to capture the process of reversion by a simple probabilistic process and derive results that are likely to be robust to the details of genetic reversion . Fourth , although we considered models of immunisation that included persistently “missed” populations , we did not explicitly consider geographic heterogeneity in coverage and risk . Areas with poor routine immunisation coverage are also often challenging places to deliver vaccine during mass campaigns . These heterogeneities in coverage are therefore likely to increase the risk of VDPV2 emergence associated with tOPV SIAs , and it may be advisable to increase the number of campaigns in areas with heterogeneous coverage to account for this risk , as has been necessary during the eradication of wild-type polioviruses . Finally , other variables that could play a role in VDPV emergence such as hygiene behaviour , sanitation or the prevalence of other enterovirus serotypes ( which could act as partners for recombination with Sabin viruses ) were not included in the analysis of data from Nigeria , because they were not available at the district level . Our transmission models do not account for the introduction of 1 or more doses of IPV into routine immunisation , which is a pre-requisite for OPV2 withdrawal [1] . Most countries currently using OPV only will introduce a single IPV dose at 14 or 16 weeks . This is likely to protect about 50% of vaccine recipients against poliomyelitis [35] . However , it is unclear what impact this vaccine will have on poliovirus transmission , which may well be limited as a result of the poor mucosal protection induced by this vaccine [36] . In particular , children born post OPV2 cessation ( thus non-exposed to live serotype-2 virus ) may not benefit from the boost in mucosal immunity induced by IPV [37 , 38] . Therefore , the impact of IPV on VDPV2 emergence and transmission may be limited [39] , leaving populations at risk of silent transmission of poliovirus . Environmental surveillance will play a major role during the polio endgame , in particular to detect highly divergent Sabin-2 viruses that could be silently circulating and help to determine the extent of possible SIAs with monovalent OPV2 to control the spread of those and avoid new VDPV2 outbreaks . Taken together , our results have important implications for efforts to prevent the emergence and subsequent spread of VDPVs during the polio endgame . First , they highlight the importance of enhancing routine immunisation coverage , which is already one of the main objectives of the Polio Eradication & Endgame Strategic Plan 2013–2018 [1] . Second , they may help to define a strategy for the use of tOPV in SIAs preceding the withdrawal of serotype 2 OPV that minimises the risk of VDPV2 post OPV2 withdrawal . In settings where population immunity is low and cVDPV2 currently absent , our findings and past experience suggest that 1 or 2 tOPV SIAs could increase the ( small ) risk of VDPV2 emergence compared to doing nothing . This risk is enhanced where tOPV immunogenicity is low , SIA coverage poor or there is a persistently “missed” group . Therefore , if tOPV SIAs are implemented preceding OPV2 withdrawal then they should be of sufficient number and high coverage to achieve high serotype-2 population immunity . In the context of limited resources , these SIAs should be targeted to countries considered at high risk of VDPV2 emergence according to the WHO risk assessment system . Finally , given the recently demonstrated advantage of IPV compared with OPV in terms of the boost to humoral and intestinal immunity in previously OPV-immunised children [37 , 38] , IPV SIAs in high-risk areas may also be considered as part of the strategy to minimise the risk of VDPV2 emergence during OPV2 withdrawal .
|
Global , coordinated withdrawal of serotype-2 OPV ( OPV2 ) is planned for April 2016 and will mark a major milestone for the Global Polio Eradication Initiative ( GPEI ) . Because OPV2 withdrawal will leave cohorts of young children susceptible to serotype-2 poliovirus , minimising the risk of new serotype-2 vaccine-derived poliovirus ( VDPV2 ) emergences before and after OPV2 withdrawal is crucial to avoid large outbreaks . Supplementary immunisation activities ( SIAs ) with trivalent OPV ( tOPV ) could raise serotype-2 immunity in advance of OPV2 withdrawal , but may also create new VDPV2 . To guide the GPEI strategy we examined the risks and benefits of implementing tOPV SIAs using mathematical models and analysis of data on the 29 independent VDPV2 emergences in Nigeria during 2004–2014 . We found that in settings with low routine immunisation coverage , the implementation of a small number of tOPV SIAs could in fact increase the probability of VDPV2 emergence . This probability is greater if SIA coverage is poor or if there are persistently unvaccinated groups within the population . A strategy of tOPV SIA in sufficient number and with high coverage to achieve high population immunity in geographically-focused , at-risk areas is needed to reduce the global risk of VDPV2 emergence after OPV2 withdrawal .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"children",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"geographical",
"locations",
"microbiology",
"mathematical",
"models",
"viruses",
"preventive",
"medicine",
"vaccines",
"age",
"groups",
"rna",
"viruses",
"vaccination",
"and",
"immunization",
"africa",
"families",
"research",
"and",
"analysis",
"methods",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"medical",
"microbiology",
"mathematical",
"and",
"statistical",
"techniques",
"microbial",
"pathogens",
"poliomyelitis",
"nigeria",
"people",
"and",
"places",
"enteroviruses",
"immunity",
"poliovirus",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"population",
"groupings",
"viral",
"diseases",
"organisms"
] |
2016
|
Preventing Vaccine-Derived Poliovirus Emergence during the Polio Endgame
|
Cellular heterogeneity , which plays an essential role in biological phenomena , such as drug resistance and migration , is considered to arise from intrinsic ( i . e . , reaction kinetics ) and extrinsic ( i . e . , protein variability ) noise in the cell . However , the mechanistic effects of these types of noise to determine the heterogeneity of signal responses have not been elucidated . Here , we report that the output of epidermal growth factor ( EGF ) signaling activity is modulated by cellular noise , particularly by extrinsic noise of particular signaling components in the pathway . We developed a mathematical model of the EGF signaling pathway incorporating regulation between extracellular signal-regulated kinase ( ERK ) and nuclear pore complex ( NPC ) , which is necessary for switch-like activation of the nuclear ERK response . As the threshold of switch-like behavior is more sensitive to perturbations than the graded response , the effect of biological noise is potentially critical for cell fate decision . Our simulation analysis indicated that extrinsic noise , but not intrinsic noise , contributes to cell-to-cell heterogeneity of nuclear ERK . In addition , we accurately estimated variations in abundance of the signal proteins between individual cells by direct comparison of experimental data with simulation results using Apparent Measurement Error ( AME ) . AME was constant regardless of whether the protein levels varied in a correlated manner , while covariation among proteins influenced cell-to-cell heterogeneity of nuclear ERK , suppressing the variation . Simulations using the estimated protein abundances showed that each protein species has different effects on cell-to-cell variation in the nuclear ERK response . In particular , variability of EGF receptor , Ras , Raf , and MEK strongly influenced cellular heterogeneity , while others did not . Overall , our results indicated that cellular heterogeneity in response to EGF is strongly driven by extrinsic noise , and that such heterogeneity results from variability of particular protein species that function as sensitive nodes , which may contribute to the pathogenesis of human diseases .
Intracellular signaling pathways must respond appropriately to various signals from the external environment . However , a variety of noise inside and outside of the cells can evoke heterogeneous responses in individual cells even when exposed to the same stimuli [1 , 2] . Although such heterogeneity interferes with a precise signaling response , it often plays essential roles in biological functions . Examples include diverse responses between amoeba cells that can undergo collective chemotaxis , and enhancement of signal entrainment in NF-κB response over a wider range of dynamic inputs by cellular noise [3 , 4] . Cellular noise is categorized into intrinsic and extrinsic noise [5–7] . Intrinsic noise is generally evoked by small numbers of molecules , representing fluctuations in biochemical reactions , transcriptional noise , molecular diffusion , etc . One of the best-known examples is the stochastic gene expression in Escherichia coli [5] . On the other hand , extrinsic noise is defined by the differences in amounts of proteins in individual cells ( protein variability ) and external physical environments , such as cell-to-cell contact , cell cycle phase and cell shape . The relationships between heterogeneous cellular responses and extrinsic noise in various signaling pathways have been reported [4 , 8–10] . However , the mechanistic effects of these types of noise to determine heterogeneity of signal responses are unclear . In this study , using mathematical modeling and simulations , we determined how cellular noise regulates heterogeneous cell responses , focusing on the epidermal growth factor ( EGF ) signaling pathway as an example . The EGF signaling pathway regulates cell growth , proliferation , differentiation , and apoptosis [11 , 12] . EGF ligands bind to EGF receptors ( EGFR ) , and the signal is transmitted to an intracellular biochemical reaction network . This signal transduction eventually phosphorylates extracellular signal-regulated kinase ( ERK ) and causes transient accumulation of ERK at the nucleus [13 , 14] . The EGF dose response of phosphorylated ERK shows a graded response [10 , 15–17] . However , it was recently reported that the dose response of nuclear ERK activity is in fact switch-like [14]; a threshold mechanism regulated by ERK may be involved in cell fate decision . The switch-like behavior is more sensitive to perturbations than the graded response [18] , and hence the effect of biological noise is considered to be critical to determine nuclear ERK activity . In fact , heterogeneous cell responses in nuclear ERK have been observed [13 , 14] . To determine how such heterogeneity in nuclear ERK response is evoked , we performed mathematical modeling and simulation analysis of the EGF signaling pathway . We developed a mathematical model of the EGF signaling pathway integrating feedback regulation between ERK and nuclear pore complex ( NPC ) , which is essential for switch-like activation of nuclear ERK , and previously developed mathematical schemes [19–21] . We also developed a new method to compare simulation results with experimental data by estimating Apparent Measurement Error ( AME ) . Finally , we elucidated how intrinsic and extrinsic noise regulate heterogeneity in nuclear ERK responses .
As shown in Fig 1 , we developed a novel mathematical model based on the biological reaction networks of the EGF signaling pathway integrated with autoregulatory control of ERK translocation . The details of our model are as follows . The reaction scheme of EGF signaling pathway is based on several published mathematical models [19–21] . EGF signaling is initiated by binding between EGF ligands and EGFR on the cell membrane , and EGF–EGFR complexes are subsequently dimerized and autophosphorylated [22–24] . Phosphorylated EGFR dimer ( pEGFR ) transmits the signal through two pathways , i . e . , the src homology and collagen protein ( Shc ) -independent/dependent pathways . Shc bound to pEGFR associates with growth factor receptor-bound protein 2 ( Grb2 ) , while Grb2 can directly associate with pEGFR [25 , 26] . Grb2 in both pathways recruits Son of Sevenless ( Sos ) from the cytoplasm to the membrane , which binds to the membrane-anchored protein Ras [27 , 28] . This association leads to exchange of guanosine diphosphate of Ras ( RasGDP ) for guanosine triphosphate ( RasGTP ) . The inactivation of RasGTP is mediated by GTPase activating protein ( GAP ) [19 , 20 , 29] . The details of this reaction scheme are shown in S1A Fig . Although little is known about the detailed reaction processes involved in Raf activation , a model of Raf activation was recently proposed based on single-molecule observations [30 , 31] . In this model , both RasGDP and RasGTP can associate with Raf . However , the association rate between Raf and RasGTP was higher than that of RasGDP . Only the RasGTP–Raf complex is able to activate Raf through an intermediate state . Kinetic parameters in the reactions were estimated from experimental data [31] . This activation scheme of Raf is included in our model ( S1B Fig ) . Activated Raf doubly phosphorylates cytoplasmic MEK ( ppMEK ) , and subsequently ppMEK also doubly phosphorylates ERK ( ppERK ) . In addition , ppERK inhibits Sos through phosphorylation , which acts as negative feedback in EGF signaling [32 , 33] . We assumed that Raf , MEK , and ERK are dephosphorylated by different phosphatases [20] . All biochemical reactions related to the EGF signaling pathway in our model are shown in S1A and S1B Fig . ERK transiently translocates into the nucleus through binding with NPC after EGF stimulation [13] . Several regulatory mechanisms of ERK translocation have also been proposed [14 , 34 , 35] . ERK-mediated phosphorylation of NPC reduces the nuclear accumulation of importin-beta that transports several proteins , including ERK , from the cytoplasm to the nucleus [34 , 35] , suggesting that activated ERK may potentially regulate its own translocation . In addition , we recently demonstrated that ERK-mediated phosphorylation of NPC is involved in the switch-like behavior of nuclear ERK translocation [14] . Based on these biological findings , we developed a new mathematical model to describe ERK translocation ( S1C Fig ) . In our model , cytoplasmic and nuclear ERK bind to NPC and translocate between the cytoplasm and the nucleus . An NPC has multiple phosphorylation sites for ERK in FG nucleoporins , which regulate the permeability barrier properties of the NPC [34 , 36] . The dynamic behaviors of such multiple phosphorylation systems have been reproduced using two-step reaction models . For example , retinoblastoma tumor suppressor protein regulated by multiple phosphorylation was modeled using two-step phosphorylation , i . e . , considering non- , hypo- , and hyperphosphorylated forms in several models [37–39] . Therefore , we introduced two phosphorylation states of NPC that are mediated by nuclear ppERK into our model ( pNPC and ppNPC in S1C Fig ) . Further , it was reported that the translocation rate of ERK was dependent on phosphorylation states of both ERK and NPC [14 , 34 , 40] . To clarify the effect of NPC phosphorylation on the ERK translocation , we assumed the following: the translocation rates of non-phosphorylated and phosphorylated ERK are different among the phosphorylation states of NPC , both NPC and pNPC mediate the translocation of phosphorylated ERK from cytoplasm to nucleus , and ppNPC allows unidirectional translocation of ERK , from nucleus to cytoplasm ( S4 Table ) . Overall , in our model , cytoplasmic ERK is phosphorylated and translocates into the nucleus transiently after EGF stimulation . Thereafter , nuclear ppERK phosphorylates NPC in a two-step process , which finally induces the export of phosphorylated ERK from the nucleus to the cytoplasm ( Fig 1 ) . Our model consists of 78 chemical species and 150 biochemical reactions . The initial conditions , i . e . , the number of molecules , of each species are shown in S1 Table . ERK and its phosphatases were considered to be distributed in both the cytoplasm and the nucleus ( Fig 1 ) . The biochemical reaction processes , association , dissociation , phosphorylation , dephosphorylation , and degradation , were described by mass-action law ( S2 Table ) . Details of the simulation method and how to determine the kinetic parameters are described in the Materials and Methods .
To confirm the biological validity of our mathematical model , we first implemented deterministic simulations . Here , pERK represents the total amount of singly/doubly phosphorylated ERK , including their complexes , and nERK represents the fold change in nuclear ERK , defined as the ratio of the total amount of nuclear ERK to the initial value . Simulated time courses of both pERK and nERK showed transient dynamics , and peak levels increased with elevated EGF concentration ( Fig 2A and 2B ) . These dynamics were consistent with the typical dynamics after EGF stimulation observed in cell lines of various types [13 , 14 , 19 , 40] . Next , we calculated the EGF dose response of peak levels of pERK and nERK ( Fig 2C and 2D ) , which showed good agreement with the experimental data [14] . The dose response of pERK showed a graded pattern ( Hill coefficient = 1 . 46 ) , while that of nERK showed switch-like behavior ( Hill coefficient = 2 . 99 ) . To confirm that this difference was caused by ERK-mediated regulation of NPC , we performed simulation without the regulation from ERK to NPC . To remove this regulation , the kinetic parameters related to ERK-mediated phosphorylation of NPC were set to zero ( reaction number 137–144 in S1C Fig , bottom right ) . As shown in Fig 2E and 2F , the dose response of nERK changed from switch-like to graded , and the Hill coefficient of nERK corresponded to that of pERK ( Hill coefficient = 1 . 46 ) . This result indicated that ERK-mediated phosphorylation of NPC is responsible for the switch-like response of nERK . The ERK-mediated phosphorylation of NPC accelerates a nuclear export of ERK in our model , establishing a negative autoregulation of nuclear ERK ( Fig 1 ) . To investigate the mechanism by which the negative autoregulation changed dynamics of nERK , the dose response of nERK at fold change level was shown in S2 Fig . The negative autoregulation of nuclear ERK drastically reduced the maximum fold change level of nuclear ERK , resulting in decreasing the range of effective concentration ( EC ) 10 and EC90 ( S2 Fig ) . While the level of EC90 was decreased from 1 . 00 to 0 . 12 by the negative autoregulation , the level of EC10 did not change . This is because ERK-mediated phosphorylation of NPC regulates the nuclear export but not the nuclear import ( Fig 1 ) . As the range of EC10 and EC90 becomes narrow , Hill coefficient is increased . Therefore , a reduction in EC90 level caused by ERK-mediated regulation of NPC enabled the dynamics of ERK translocation to be changed from graded to switch-like . Indeed , knockdown of nucleoporin 153 , one of the relevant components of NPC that is most effectively phosphorylated by ERK , altered the dose response of nuclear ERK from switch-like to graded [14] . Thus , our model could recapitulate the essential dynamics of the EGF signaling pathways , suggesting that our model can be used for further simulation analysis . To investigate the effects of intrinsic and extrinsic noise on heterogeneity in nuclear ERK activity , we implemented simulations with either intrinsic or extrinsic noise ( see Materials and Methods for details ) . In this study , the intrinsic and extrinsic noise were defined as fluctuation in reactions and protein variability , respectively . Here , fluctuation in the reactions means that biochemical reaction occurs stochastically , which can be simulated using the Gillespie algorithm [41] . On the other hand , protein variability means that there are differences in the levels of proteins between individual cells , and the noise level is represented by the coefficient of variation ( CV ) . In these simulations , we used a typical CV value of protein variability , 30% , as a representative value of extrinsic noise [8 , 42] . The distributions of peak levels of nuclear ERK obtained from simulations with intrinsic or extrinsic noise are shown in Fig 3 . The distribution with extrinsic noise was clearly broader than that with intrinsic noise at high concentrations of EGF ( Fig 3A ) . For statistical comparison , the CV of nuclear ERK was calculated from simulated distributions . The CV of nuclear ERK with extrinsic noise was higher than that with intrinsic noise ( Fig 3B ) , suggesting that extrinsic but not intrinsic noise contributed to the heterogeneity in nuclear ERK activity . The variability of proteins between individual cells can be measured by various experimental methods . However , it is still difficult to measure variability of all protein species present in a mammalian cell . To estimate all protein variability in the EGF signaling pathway , we directly compared simulations with experiments . In addition to biological noise , the observed data included several measurement errors derived from measurement principles and setups . Here , such measurement errors were defined as the Apparent Measurement Error ( AME ) , which was determined by our newly developed method ( details are described in Supporting Information ) . As shown in S3 Fig , by applying AME , the distribution of nuclear ERK in simulations corresponded to the observed data [14] . Using the identified AME , we estimated the variability of all proteins in the EGF signaling pathway . First , simulations with both types of noise were implemented under different concentrations of EGF when the CV of protein variability changed from 0% to 50% . The resulting distributions of fold changes in nuclear ERK without and with AME are shown in S4 and S5 Figs , respectively . The CV of nuclear ERK response was calculated for statistical comparison of these simulation results with experimental data [14] . As shown in Fig 4A , AME strongly affected the distributions at low EGF concentration ( < 0 . 01 ng/mL ) but had little effect at high concentrations , and by applying AME the CV of nuclear ERK corresponded to the pattern of experimental data . Simulation results at 25% CV of protein variability showed excellent agreement with experimental data ( Fig 4A and 4B ) . For further quantitative comparison , we calculated mutual information , which has been proposed as a good metric to characterize fidelity for a biological system [43] . As shown in Fig 4C , the mutual information between EGF and nuclear ERK also showed that 25% CV of protein variability was the best fit to the experimental value [14] . The distributions of nuclear ERK in simulation results at 25% CV of protein variability were also consistent with experimental data ( Fig 4D ) . Thus , our new method using AME made it possible to predict variability of signaling proteins from only signal output data . In our simulations , extrinsic noise was reproduced by sampling initial proteins randomly from a log-normal distribution . However , it has been reported that individual cells have different expression capacities , leading to variations in the levels of proteins in a correlated manner [6] . To investigate the effects of such covariation among proteins on heterogeneity , we implemented simulations in which all protein levels under the initial conditions were correlated ( Fig 5A ) . The covariation among proteins did not influence the distribution of nuclear ERK at the steady-state level without EGF stimulation , as shown in Fig 5B . However , at high concentrations of EGF , CV of nuclear ERK with covariation was lower than that with covariation ( Fig 5C ) . This tendency for covariation to suppress the heterogeneity in nuclear ERK was found regardless of the application of AME ( Fig 5C ) . Thus , our simulation results suggested that covariation among proteins is involved in heterogeneous cellular responses in the EGF signaling pathway . To investigate the contribution of each molecular species to heterogeneity in nuclear ERK , we implemented simulations by changing the variability of each protein . The effects of the variability of each protein on heterogeneity in nuclear ERK at low ( 0 . 05 ng/mL ) and high ( 50 ng/mL ) concentrations of EGF are shown in Fig 6A and 6B , respectively . At lower EGF , variability of EGFR , Ras , Raf , and MEK generated marked heterogeneity in nuclear ERK , whereas variability of ERK and Sos generated large degrees of heterogeneity at higher EGF concentrations . These results suggest that different proteins contributed to cellular heterogeneity in nuclear ERK at different EGF concentrations . Therefore , we investigated the contributions of the proteins in the presence of various concentrations of EGF ( Fig 6C ) . The contributions to heterogeneity in nuclear ERK were divided into the following three types: ( 1 ) EGFR , Ras , Raf , and MEK; ( 2 ) ERK and Sos; ( 3 ) GAP , Grb2 , and Shc . Species in the first type evoked large heterogeneity between effective concentration ( EC ) 10 and EC90 ( Fig 6C , top ) , where cells that did and did not respond to EGF stimulus were mixed ( Fig 4D ) . Accordingly , heterogeneity generated by variability of EGFR , Ras , Raf , and MEK would be closely related to the response to EGF stimulus . On the other hand , species in the second type stably generated large degrees of heterogeneity over EC90 ( Fig 6C , middle ) . In this region , heterogeneous responses occurred at high levels of nuclear ERK , indicating that all cells would respond to the stimulus ( Fig 4D ) . Therefore , variability of ERK and Sos had little effect on the response to EGF . Species in the third type showed little heterogeneity at any concentration of EGF ( Fig 6C , bottom ) , and therefore these species had no contribution to the response to EGF in addition to those in type two . This result indicated that only particular species involved in the EGF signaling pathway , i . e . , EGFR , Ras , Raf , and MEK , regulate heterogeneity of nuclear ERK in relation to EGF signaling response , i . e . , these proteins function as sensitive nodes in the signaling response . In the apoptotic pathway , whether apoptosis was induced or not was not correlated with variability of any single species included in the pathway [8] , suggesting that heterogeneity was regulated by variability of at least more than two species . Our simulation results indicated the possibility of predicting EGF signaling response at the single-cell level by knowing the initial concentrations or variability of four species , i . e . , EGFR , Ras , Raf , and MEK .
In this study , we developed a novel mathematical model of the EGF signaling pathway integrated with the mechanisms regulating the nuclear translocation of ERK . Although the nuclear translocation of ERK is critical for cell fate decision [44] , the dynamics and the regulation mechanism have not been taken into consideration in conventional mathematical models [19–21] . Our model included ERK-mediated regulation of NPC explicitly , which could realize the observed dynamics of nuclear ERK , i . e . , switch-like behavior . Our model assumed that nuclear ppERK phosphorylates the NPC in a two-step process , and then ppNPC but not NPC and pNPC mediated the translocation of nuclear phosphorylated ERK to the cytoplasm . The nuclear ppERK positively regulates its own nuclear export through NPC , and therefore activated ERK inhibits its own accumulation in the nucleus , which generated a negative autoregulation of nuclear ERK . Without this negative autoregulation , as phosphorylated ERK was simply distributed in both the cytoplasm and the nucleus through NPC , nERK and pERK showed the same dynamics , i . e . , graded response . Thus , ERK-mediated regulation of NPC was responsible for the switch-like response , which may play a crucial role in cell fate decision . Although the details of the molecular basis underlying ERK nuclear translocation are still controversial and our model includes several assumptions regarding the regulatory mechanism between ERK and NPC , we emphasize that our model captures the essential behaviors of ERK in response to EGF stimulation , including time course and dose response . Next , we investigated the effects of intrinsic and extrinsic noise , i . e . , fluctuations in reactions and protein variability , on heterogeneity in nuclear ERK and found that extrinsic rather than intrinsic noise contributed to cellular heterogeneity . Our model assumed the EGF signaling pathway in a mammalian cell in which the volume is typically > 10−12 L , which is much larger than yeast or bacteria ( 10−16–10−14 L ) [45 , 46] . Therefore , a mammalian cell has a huge number of molecules even if the same concentrations of proteins are present in yeast and bacterial cells . For example , the EGF signaling pathway consists of 103–107 molecules [40] and the HGF signaling pathway consists of 104–107 molecules [9] . On the other hand , E . coli and yeast cells possess 10−1–103 and 102–106 protein molecules , respectively [47 , 48] . In general , fluctuations due to intrinsic noise are strongly dependent on the number of molecules , i . e . , fluctuations are larger with smaller numbers of molecules . Therefore , intrinsic fluctuation is considered to be negligibly small , and extrinsic noise evokes larger fluctuations in our model . In addition , it was reported that the abundance of transcripts in mammalian cells was regulated by extrinsic noise , i . e . , cellular state , population context , and microenvironment [49] , indicating that extrinsic noise plays a key role in the transcriptional program . Our simulations demonstrated the importance of extrinsic noise in the signaling response . These results suggest that extrinsic noise plays significant roles in mammalian cellular responses . Using a newly developed method to estimate AME , we could predict that CV of protein variability in the EGF signaling pathway was 25% . In other signaling pathways , ranges of measured CV of signaling proteins were 8%– 75% in the hepatocyte growth factor ( HGF ) signaling pathway [9] , 21%– 28% in the apoptotic intrinsic pathway [8] , and 15%– 30% in normally cycling human cells [42] . Thus , estimated CV was included in the range of observed protein variability , indicating that our method is useful for estimating or predicting the variability of all signaling proteins . The AME estimated by our method agreed completely with observed heterogeneity in nuclear ERK without EGF stimulation , suggesting that at the basal level , variability arising from cellular noise was too small to contribute to cellular heterogeneity . Moreover , in our simulations , covariation among proteins under initial conditions suppressed variation in nuclear ERK at high concentrations of EGF . This suggests that heterogeneous signaling responses can be regulated by such covariation in the cell . Further analysis using estimated protein variability showed that distinct species in EGF signaling pathway have different effects on heterogeneity in nuclear ERK , i . e . , EGFR , Ras , Raf , and MEK generated heterogeneity related to the signaling response . These particular proteins function as sensitive nodes causing heterogeneous cell responses in the EGF signaling pathway . On the other hand , proteins other than sensitive nodes , i . e . , GAP , Grb2 , and Shc , could not influence the heterogeneity at any concentrations of EGF . Such differential contribution to heterogeneity may be due to the mechanism of reactions involving each protein in the pathway . In our model , sensitive nodes are involved in enzymatic reactions such as phosphorylation , while insensitive nodes are related to binding–unbinding reaction . This suggests that heterogeneity in signaling responses may be regulated by the type of network edges , i . e . , reaction mode in the signaling pathway . In terms of cellular functions , the expression level or the activity of sensitive nodes would be tightly regulated in the cell for appropriate responses in the signaling pathway , as they are strongly involved in the cellular heterogeneous responses . In fact , mutations of EGFR , Ras , and Raf are related to epithelial mesenchymal transition , migration , and tumor invasion of breast cancers , and furthermore MEK mutation was observed in malignant melanoma [50–52] . Thus , sensitive nodes seem to be committed to maintain normal cellular homeostasis . As the expression levels of these sensitive nodes are closely related to the signaling response , knowing these concentrations may make it possible to predict the signaling response at the single-cell level before stimulation .
Deterministic and stochastic simulations were implemented in this study . Two types of cellular noise , i . e . , intrinsic and extrinsic noise , were simulated . The intrinsic noise was defined by fluctuations in reactions , which were realized by a stochastic simulation method ( Gillespie algorithm ) [41 , 53] . The extrinsic noise was defined by protein variability between individual cells , which was represented by independently sampling initial values of each protein from log-normal distributed random variables at various CV [8 , 9 , 18] . We implemented four types of simulation: 1 ) without either type of noise; 2 ) with only intrinsic noise; 3 ) with only extrinsic noise; 4 ) with both types of noise . All kinetic parameters were constant among all simulations except the CV of protein variability ( S1 , S2 and S3 Tables ) . In the case of stochastic simulations , we performed 5000 simulations per condition to obtain statistically stable results . The initial conditions of five species , EGFR , Ras , Raf , MEK and ERK , were estimated from experimental data [22 , 40] . Similarly , six kinetic parameters related to Raf activation were experimentally determined values ( S2 Table ) . Other initial conditions and kinetic parameters were determined by manual parameter tuning as follows . First , kinetic parameters in published models were used as the initial estimates [20 , 40] . Then , we performed deterministic simulations changing each parameter one by one , and compared the simulation results with experimental data , i . e . , time course and dose response of both phosphorylated and nuclear ERK [14] . Repeating this process , we finally determined the parameter set that reproduced the experimental data . The mutual information between EGF stimulus and nuclear ERK response was calculated from simulation results and experimental data . Here , the concentration of EGF and nuclear ERK level were used as input signal ( S ) and output response ( R ) , respectively . Mutual information , I ( R; S ) , was defined by the following equation: I ( R;S ) =ΣsΣRP ( R , S ) log2 ( P ( R , S ) P ( R ) P ( S ) ) ( 1 ) where P ( R , S ) , P ( R ) , and P ( S ) represent the joint probability distribution functions of S and R , and the marginal probability distribution functions of S and R , respectively . As the distributions of simulated S and R were discretized , direct estimates of mutual information using ( Eq 1 ) were biased . To obtain the unbiased solution , we calculated the mutual information by optimizing the discretized size and jackknife sampling , as described previously [54] . The experimental data used in this research were originally reported in our previous paper [14] .
|
Individual cell behaviors are controlled by a variety of noise , such as fluctuations in biochemical reactions , protein variability , molecular diffusion , transcriptional noise , cell-to-cell contact , temperature , and pH . Such cellular noise often interferes with signal responses from external stimuli , and such heterogeneity functions in induction of drug resistance , survival , and migration of cells . Thus , heterogeneous cellular responses have positive and negative roles . However , the regulatory mechanisms that produce cellular heterogeneity are unclear . By mathematical modeling and simulations , we investigated how heterogeneous signaling responses are evoked in the EGF signaling pathway and influence the switch-like activation of nuclear ERK . This study demonstrated that cellular heterogeneity of the EGF signaling response is evoked by cell-to-cell variation of particular signaling proteins , such as EGFR , Ras , Raf , and MEK , which act as sensitive nodes in the pathway . These results suggest that signaling responses in individual cells can be predicted from the levels of proteins of sensitive nodes . This study also suggested that proteins of sensitive nodes may serve as cell survival mechanisms .
|
[
"Abstract",
"Introduction",
"Model",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"phosphorylation",
"egfr",
"signaling",
"mathematical",
"models",
"simulation",
"and",
"modeling",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"erk",
"signaling",
"cascade",
"proteins",
"mathematical",
"and",
"statistical",
"techniques",
"ras",
"signaling",
"cytoplasm",
"biochemistry",
"biochemical",
"simulations",
"signal",
"transduction",
"cell",
"biology",
"post-translational",
"modification",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"cell",
"signaling",
"signaling",
"cascades"
] |
2016
|
Modeling Cellular Noise Underlying Heterogeneous Cell Responses in the Epidermal Growth Factor Signaling Pathway
|
Association testing of multiple correlated phenotypes offers better power than univariate analysis of single traits . We analyzed 6 , 600 individuals from two population-based cohorts with both genome-wide SNP data and serum metabolomic profiles . From the observed correlation structure of 130 metabolites measured by nuclear magnetic resonance , we identified 11 metabolic networks and performed a multivariate genome-wide association analysis . We identified 34 genomic loci at genome-wide significance , of which 7 are novel . In comparison to univariate tests , multivariate association analysis identified nearly twice as many significant associations in total . Multi-tissue gene expression studies identified variants in our top loci , SERPINA1 and AQP9 , as eQTLs and showed that SERPINA1 and AQP9 expression in human blood was associated with metabolites from their corresponding metabolic networks . Finally , liver expression of AQP9 was associated with atherosclerotic lesion area in mice , and in human arterial tissue both SERPINA1 and AQP9 were shown to be upregulated ( 6 . 3-fold and 4 . 6-fold , respectively ) in atherosclerotic plaques . Our study illustrates the power of multi-phenotype GWAS and highlights candidate genes for atherosclerosis .
Five years of genome-wide association studies ( GWAS ) have successfully identified common variants at >1 , 000 genomic loci robustly associated with a wide range of human conditions and quantitative traits [1] . Despite this progress , one limitation is that almost all GWAS performed to date have focused on single traits , even in studies involving multiple related phenotypes . Growing evidence for pleiotropy [2] , [3] , where the same locus is associated with multiple phenotypes , supports the idea that multivariate analysis of multiple phenotypes can provide a substantial boost in power for locus discovery , consistent with simulation studies [4]–[7] . A plethora of metabolites in blood have been described as risk factors for metabolic syndrome , atherosclerosis and coronary artery disease [8] , [9] . Using high-throughput nuclear magnetic resonance assays , quantitative profiles of 130 metabolites in two population-based cohorts from Finland , the Cardiovascular Risk in Young Finns Study ( YFS ) [10] and the Northern Finland Birth Cohort 1966 ( NFBC66 ) [11] have been determined . These metabolites included lipoprotein subclasses of VLDL , LDL , IDL and HDL as well as lipids , amino acids and other small molecules ( Materials S1 ) . Figure 1 illustrates the general flow of our study . We first applied an unsupervised algorithm to identify networks from the observed correlation structure amongst the 130 metabolite measures in 6 , 600 individuals . For each of these networks , we performed a multivariate test of association for 2 . 5 million SNPs [4] . Because we also tested all SNPs for association to each metabolite separately , we can assess the relative gain in power of the multivariate approach . To interpret the novel signals , we tested whether the associated SNPs influenced cis-gene expression levels in multiple tissues as well as whether the expression of candidate genes was associated to specific metabolites that drive the initial association . Finally , we analysed arterial tissue from mouse and man to test for a relation between our top candidate genes and atherosclerosis plaques .
We analysed genotype and phenotype data from the YFS ( N = 1 , 905 ) [10] , [12] and the NFBC66 ( N = 4 , 703 ) [11] , [13] . For both YFS and NFBC66 , we imputed SNP genotypes using the MACH algorithm [14] and the HapMap Phase 2 reference panel [15] . Serum collected from both cohorts underwent metabolomic profiling on the same proton nuclear magnetic resonance ( NMR ) platform [16] . The NMR metabolomics platform used here provided absolute quantitative information on 130 distinct metabolic measures [17] . Metabolite levels for both cohorts were normalized and adjusted for age , gender , cohort , and population structure ( Materials and Methods ) . After correcting for cohort effects and pooling the metabolomic data for YFS and NFBC66 , we constructed a Pearson correlation matrix that defined the pairwise relationships between all metabolites and applied agglomerative hierarchical clustering in order to identify networks of metabolites ( Figure 2 ) . Using a dynamic , data-driven tree-cutting algorithm [18] , we identified 11 metabolic networks that represent various metabolic pathways ( Materials and Methods and Figure S1 ) . Additional information for each metabolic network , including full descriptions , abbreviations , inter-metabolite correlations , and supporting association analyses , is given in the Materials S1 . Briefly , metabolic network 1 comprises multiple measures related mainly to cholesterol metabolism in the apoB-containing lipoproteins . Metabolic network 2 includes branched-chain and aromatic amino acids together with the large TG-rich VLDL particles and serum triglycerides . Metabolic networks 3 and 4 capture the larger and smaller particles of HDL-metabolism , respectively . Metabolic networks 5 , 6 , 7 , and 8 are related to lipid poly-unsaturation , ketone bodies , the glucose-alanine cycle , and renal function , respectively . Metabolic networks 9 , 10 , and 11 each contain only 2 metabolites and represent measures of fatty acid chain length and composition , mean diameter of LDL and double-bonding of fatty acid chains , urea and acetate , respectively . For each of the 11 metabolic networks , we performed SNP association testing using a multivariate test based on Canonical Correlation Analysis and Wilks' lambda [4] . Each association test yielded an F statistic , corresponding P value , and a loading for each metabolite in the network to indicate the relative contribution of that metabolite to the overall association ( Materials and Methods ) . For univariate analysis , we used standard linear regression where each of the 130 metabolites was regressed onto each SNP . The implementation of dimensionality reduction and multivariate analysis allowed us to select essentially independent tests based on the correlation structure of the phenotype data . Using multivariate analysis , we tested each SNP only 11 times ( one per metabolic network ) . A Bonferroni correction for testing each SNP to 130 metabolites is overly conservative , since the metabolites are partially correlated , but still common practice [16] , [19] , [20] . Accordingly , we set genome-wide significance thresholds at P<4 . 5×10−9 and P<3 . 8×10−10 for multivariate and univariate analysis , respectively . To maximize power , we performed a joint analysis of both cohorts , correcting for population structure and cohort-specific effects . We observed little evidence for test statistic inflation , lambda range 1 . 01–1 . 06 ( Materials and Methods and Table S1 ) . Across all 11 metabolic networks , the joint multivariate analysis yielded 713 SNPs significantly associated with one or more metabolic networks ( P<4 . 5×10−9 ) . This corresponded to 34 distinct loci and 75 significant locus-network associations overall ( Table 1 and Materials and Methods ) . Loci were considered novel if they had not been previously associated at genome-wide significance with a metabolite or other metabolic phenotype in the NHGRI Catalogue of Published GWAS [1] and if they were independent of other proximal signals ( Materials and Methods and Table S2 ) . Of the 34 loci detected , 27 were previously identified to be associated with fasting glucose levels [21] , [22] , total measures of LDL , HDL and triglycerides [23] , [24] , bradykinin [25] , glutamine [16] , [25] , alanine-valine ratio [16] , phenylalanine [16] , citrate [16] , and sphingolipids [26] . Overall , we found 7 novel loci associated with 12 metabolic networks in total ( Table 1 ) . In comparing multivariate and univariate P values for a given SNP , we selected the lowest univariate P value for any single metabolite in a given network . We found that multivariate tests yielded more significant P values , reflecting increased power compared to univariate tests ( Figure S2 ) . In terms of number of significant associations , multivariate analysis outperformed univariate in both cohorts . When their respective genome-wide significance thresholds are applied , multivariate analysis uncovered 75 locus-metabolic network associations , whereas univariate analysis found only 40 ( almost all of which were detected by multivariate analysis ) , leading to the detection of 7 novel loci instead of one ( Figure 3 ) . This demonstrates the relative gain of multivariate testing compared to univariate testing . Notably , multivariate analysis still uncovered more associations ( 69% more ) than univariate analysis even when applying the more stringent genome-wide threshold for 130 independent metabolites . Multivariate also outperformed univariate when assessing only known loci , i . e . those with prior genome-wide significant association to metabolites in the NHGRI Catalogue of Published GWAS [1] ( Figure S3 ) . From the multivariate analysis , our strongest association signal overall was due to a nonsynonymous SNP , rs1303 ( Glu400Asp ) located in the last exon of SERPINA1 . This variant was associated with metabolic networks 1 and 2 ( P = 5 . 4×10−48 and P = 7 . 4×10−22 , respectively; Figure S4 ) . To explore the extent to which rs1303 perturbs protein structure , we utilized the PolyPhen2 algorithm [27] . PolyPhen2 predicted the Glu to Asp mutation to be benign ( naïve Bayes posterior probability = 0 . 0 and 0 . 005 for the HumDiv and HumVar training sets , respectively ) . The next strongest signal overall was an intronic SNP ( rs16939881 ) at the AQP9 locus , associated with metabolic networks 1 , 2 , 3 , and 4 ( P = 2 . 9×10−27 , P = 4 . 9×10−15 , P = 2 . 3×10−18 and P = 2 . 0×10−14 , respectively; Figure S4 ) . The metabolic network associations at AQP9 remained highly significant after conditioning on the previously identified LIPC locus , 250 Kb downstream ( Table S2 ) . We focus on our two top signals for subsequent in-depth analyses . Because our top signals are within the SERPINA1 and AQP9 genes , we assume these to be the most likely candidate genes . Using the 1000 Genomes Phase I integrated variant set and the IMPUTE2 algorithm [28] , we generated denser maps of genetic variants for the SERPINA1 and AQP9 loci . We then performed a multivariate test for each SNP with metabolic networks 1–4 as well as a conditional analysis to ascertain any independent signals in each region . After 1000 Genomes imputation , rs1303 remained the top signal at the SERPINA1 locus for metabolic networks 1 and 2 . Conditioning on rs1303 revealed an independent association between another nsSNP , rs28929474 , and metabolic networks 1 and 2 ( P = 1 . 7×10−19 and P = 3 . 7×10−13 respectively ) ( Figure S5 ) . Rs28929474 ( Glu366Lys ) lies in the last exon of SERPINA1 and , unlike rs1303 , it was predicted by PolyPhen2 to be a probable damaging mutation with a naïve Bayes posterior probability = 1 . 0 for both HumDiv and HumVar . Imputation of the AQP9 locus with the 1000 Genomes panel yielded less confidently inferred genotypes than the HapMap2 panel at the top SNP rs16939881 ( posterior probability >0 . 90 for a genotype call ) . Consequently we had less power at rs16939881 , however it still remained significantly associated with metabolic networks 1 , 2 , and 3 . Even with reduced power , conditional analysis showed that the signal at AQP9 could be explained by rs16939881 alone ( Figure S6 ) . We next investigated metabolic network associated variants for eQTL effects on SERPINA1 and AQP9 . We used three resources ( a ) the DILGOM cohort , a Finnish population-based cohort ( N = 518 ) with gene expression data ( from whole blood ) and serum metabolomic data [29] , ( b ) a subset of the EUFAM study ( N = 54 ) with familial low HDL cholesterol phenotype [30] and subcutaneous adipose tissue gene expression data , and ( c ) the Human Liver Cohort , a Caucasian cohort ( HLC; N = 178 ) with liver tissue gene expression data [31]–[33] . These three resources also comprise genome-wide SNP data . We summarize the eQTL analyses in Table S3 . In the DILGOM study , the SNP explaining the most variance in SERPINA1 expression ( rs11628917; linear regression P = 6 . 0×10−10; adjusted R2 = 0 . 07 ) was also strongly associated with metabolic networks 1 and 2 from the YFS and NFBC66 joint analysis ( P = 9 . 6×10−14 and P = 1 . 9×10−11 ) ( Figure 4 ) . In our data , there was moderate linkage disequilibrium ( LD; r2 = 0 . 47; D′ = 0 . 99 ) between the top SERPINA1 SNP ( rs1303 ) and the blood eQTL ( rs11628917 ) . Conditional analysis showed that the association of rs11628917 with both metabolic networks could be explained by rs1303 , suggesting non-independence . Rs1303 itself was nominally associated with SERPINA1 expression ( P = 0 . 01 ) . The independent signal at rs28929474 showed no evidence of influencing SERPINA1 expression ( and was not in LD with any eQTLs ) , suggesting that its primary effect may be protein structure destabilization . No blood eQTLs were detected for AQP9 . In the EUFAM study , the top and bottom 10th percentiles of HDL-C concentrations ( Finnish population age and sex specific percentiles ) were used to define high and low HDL-C groups ( N = 19 and N = 35 , respectively ) . First , we tested for differences in AQP9 and SERPINA1 between high and low HDL-C groups . Both AQP9 and SERPINA1 expression were upregulated in adipose tissue of individuals with low HDL-C ( fold changes 3 . 47 and 2 . 29 , P = 9 . 0×10−4 and P = 0 . 03 , respectively ) . Analysis of genetic variants did not yield any eQTLs at the AQP9 the locus . Given the independent but proximal signals at AQP9 and LIPC , we did detect an eQTL 210 Kb downstream within the LIPC locus that influenced the adipose expression of AQP9 ( rs1825955; P = 4 . 8×10−3 ) but not LIPC . There was low LD between rs1825955 and the top multivariate AQP9 SNP rs16939881 ( r2 = 0 . 17; D′ = 0 . 94 ) . SERPINA1 also did not harbour adipose eQTLs ( including rs11628917 , P>0 . 05 ) , indicating either potentially tissue-specific function of the SNP or lack of statistical power . The HLC allowed for the analysis of gene expression in the human liver . In the HLC , we detected eQTLs for both SERPINA1 and AQP9 ( Figure 4 ) . An eQTL in the promoter region of SERPINA1 explained 3 . 9% of the liver expression of the gene ( rs1884549; P = 4 . 3×10−3 ) . Rs1884549 was also associated with metabolic network 1 ( P = 9 . 6×10−22 ) and in moderate LD ( r2 = 0 . 38; D′ = 0 . 99 ) with rs1303 . A variant within AQP9 was associated with its expression in the liver ( rs16953360; P = 4 . 6×10−3; adjusted R2 = 0 . 04 ) as well as metabolic networks 1–4 ( P = 1 . 0×10−25; P = 4 . 7×10−14; P = 8 . 1×10−18; P = 7 . 1×10−14 respectively ) . SNPs rs16953360 and rs16939881 are in very strong LD ( r2 = 0 . 97; D′ = 0 . 98 ) . We next investigated whether there was a relationship between SERPINA1 and AQP9 and metabolites levels in the DILGOM cohort . To do this , we considered those metabolic networks associated with SERPINA1 and AQP9 SNPs then regressed individual metabolite levels on gene expression ( Table S4 ) . Genetic variation in AQP9 was associated with metabolic networks 1–4 and here we observed significant association between expression of AQP9 and two metabolites from network 1 ( XL-HDL-TG: P = 8 . 5×10−9; MobCH3: P = 7 . 2×10−5 ) . SERPINA1 harboured genetic variants associated with metabolic networks 1 and 2 , and expression of SERPINA1 was associated with eight metabolites , four from metabolic network 1 and four from metabolic network 2 ( Table S4 ) . Since genetic variation and gene expression of SERPINA1 and AQP9 were associated with lipoprotein levels , lipid transporters central to atherosclerosis , we investigated the relationship between these genes and atherosclerosis . We first investigated a mouse model ( BxH-ApoE , N = 298 ) on a hyperlipidemic apolipoprotein-E ( ApoE ) null background with liver gene expression profiles and quantified aortic lesions [34]–[36] . BxH-ApoE consisted of an F2 population derived from a backcross of mice highly susceptible to atherosclerosis ( C57BL/6J ApoE−/− ) and highly resistant ( C3H/HeJ ApoE−/− ) . The F2 population was then fed on a high-fat , western diet for 16 weeks then euthanized at 24 weeks . Using linear regression , we tested for association between liver expression of Serpina1a ( the mouse ortholog of SERPINA1 ) and AQP9 and the area of atherosclerotic lesion in the aorta . Expression of AQP9 showed significant association with atherosclerotic plaque area ( P = 5 . 0×10−3; Figure 4 ) , with samples in the top decile of AQP9 expression having on average 29% larger lesion area than those in the bottom decile . The association remained significant after correction for gender , total cholesterol , triglycerides and HDL . On this background , Serpina1a expression did not show association with lesion area ( P = 0 . 58 ) . Finally , we utilized the Tampere Vascular Study ( TVS ) a collection of atherosclerotic plaque samples from patients undergoing peripheral vascular surgery ( carotid and femoral endarterectomy and aortic bypass procedures due to atherosclerosis ) and control samples from individuals undergoing coronary artery by-pass surgery ( Materials and Methods ) . In TVS , both SERPINA1 and AQP9 showed strong association with lesion status ( Figure 4 ) . AQP9 was expressed at a 4 . 67 fold higher level in lesions compared to controls ( Mann Whitney P = 4 . 64×10−12 ) , and similarly SERPINA1 exhibited 6 . 33 fold higher expression ( Mann Whitney P = 2 . 49×10−13 ) . The TVS results suggest that both AQP9 and SERPINA1 are candidate genes for atherosclerosis .
We have empirically demonstrated the power of multivariate association testing of metabolite networks . We detected 7 novel loci and investigated the gene expression of our top loci , SERPINA1 and AQP9 , in multiple human tissues as well as their potential role in atherosclerosis . SERPINA1 was associated with metabolic networks 1 and 2 ( top metabolites: total cholesterol in IDL and mean diameter of VLDL , respectively ) , which are mainly related to cholesterol and triglyceride pathways of apoB-containing lipoproteins as well as diabetes associated amino acids [37] . SERPINA1 encodes alpha 1-antitrypsin ( A1AT ) , a protease inhibitor that protects surrounding tissues at sites of inflammation , and various studies have suggested A1AT's role in atherosclerosis . A1AT has been detected within HDL particles but not LDL [38] , although complexes of A1AT and LDL have been found in the intimal arterial wall and in human atherosclerotic lesions in the coronary artery [39] . Proteolytic degradation of LDL by murine peritoneal macrophages has been shown to be enhanced by A1AT binding , and immunostaining and in situ hybridization have also suggested that A1AT is produced by macrophages in the arterial wall [39] . AQP9 encodes aquaporin 9 , a liver glycerol channel [40] , and contains variants which showed association with metabolic networks 1 and 2 ( top metabolites: triglycerides in very large HDL and mean diameter of VLDL ) as well as networks 3 and 4 ( top metabolites: mean diameter of HDL and phosphatidylcholine ) . The proximity of AQP9 to the well-known LIPC gene 250 Kb downstream raises the question of whether the AQP9 and LIPC loci harbour independent effects . Our conditional analyses of metabolic network associated SNPs indicate that these are indeed independent genetic signals . In addition , LIPC expression in whole blood from DILGOM was not associated with metabolites from the relevant networks 1 , 2 , 3 , or 4 , and LIPC liver expression in mouse only slightly attenuated the association of AQP9 with atherosclerotic lesion area in a linear model ( P = 0 . 059 ) . In human aorta , LIPC was nominally differentially expressed between healthy and plaque samples ( P = 0 . 01 ) and did not affect the substantially larger aortic differential expression of AQP9 . Previous experiments have shown AQP9's involvement in gluconeogenesis . AQP9 mRNA and protein have been shown to be greater in human obese T2D patients relative to lean normoglycemics in adipose tissue [41] . The opposite is true in liver , suggesting that reduction in glycerol influx in hepatocytes via AQP9 could prevent excessive lipid accumulation and may reduce hyperglycaemia in obesity [41] . Further , AQP9−/− mice have previously been shown to have elevated levels of plasma glycerol and triglycerides , and inhibition of AQP9 by a small molecule inhibitor showed that it is required for glycerol-dependent glucose production in murine hepatocytes [42] , [43] . Our findings for SERPINA1 and AQP9 are consistent with the above studies suggesting associations with cardiometabolic risk factors and show that ( a ) common variants in both are associated with metabolic networks , ( b ) these variants modulate gene expression and suggest that there may be potential heterogeneous genetic control in different tissues , ( c ) expression of both genes was associated with metabolites from the relevant networks , and finally ( d ) gene expression was positively associated with atherosclerotic lesion area in mice ( AQP9 ) and upregulated in atherosclerotic tissue in humans ( SERPINA1 and AQP9 ) . We also speculate that the roles of SERPINA1 and AQP9 in atherosclerosis are tissue-specific where AQP9 displays an effect in both liver and arterial tissue and SERPINA1 only in the latter . Of the five other novel loci , there were variants proximal to ZFHX3 on Chr 16 , MYO1E on Chr 15 , as well as three independent signals at 4q13 . ZFHX3 encodes ATBF1 , a transcription factor involved in neuronal differentiation and survival [44] , [45] that has also been previously implicated in Kawasaki disease , atrial fibrillation and ischemic stroke [46] , [47] . Variants at the ZFHX3 locus were associated with metabolic network 2 where the metabolite with the greatest loading was tyrosine . Little is known about the role of the tyrosine in circulation , however a recent study [37] investigating the predictive ability of five amino acids for type 2 diabetes onset suggested that amino acid metabolism , including tyrosine , plays a role in the pathophysiology of metabolic syndrome , where it is known that individuals with either metabolic syndrome and/or diabetes are at increased risk for stroke . At 4q13 , a band that contains the ALB albumin gene , metabolic networks 1 and 4 were associated with variants 30 Kb upstream of group-specific component , a vitamin D binding protein ( top metabolites: triglycerides in IDL and albumin , respectively ) . Metabolic network 4 , with albumin as the top metabolite , was also associated with variants 10 Kb upstream of EREG and independent variants 8 Kb upstream of CXCL5 . EREG encodes epiregulin , part of a family of epidermal growth factors for which there is evidence that osmotic pressure has a role in signal transduction [48] , and CXCL5 encodes a cytokine that has previously been linked with obesity and insulin resistance [49] . Finally , 15q22 harboured intronic variants within the MYO1E gene , a non-muscle class I myosin protein , associated with metabolic network 2 ( top metabolite: total triglycerides ) . Myosin 1E has previously been shown to bind phospholipids [50] , regulate podocyte function and glomerular filtration [51] , as well as contain nsSNPs which display linkage to kidney disease [52] . This study illustrates the importance of accounting for fine-scale phenotypic structure . Although the current GWAS paradigm is based on the testing of one phenotype and one marker at a time , the quantitative phenotype profiles of individuals and corresponding biological samples are rapidly expanding in scope and depth . We are being faced with more complex multivariate phenotypic information , and biologically heterogeneous phenotypes can now be fine-mapped to reveal more informative patterns of association . Powerful statistical approaches that leverage the network covariance can provide novel insights and link genetics with disease .
The Cardiovascular Risk in Young Finns Study ( YFS ) is a population based prospective cohort study . It was conducted at 5 university departments of medical schools in Finland ( i . e . Turku , Helsinki , Kuopio , Tampere and Oulu ) , with the aim of studying the levels of cardiovascular risk factors in children and adolescents in different parts of the country . The latest follow-up was conducted in 2007 . The serum samples for this metabolomics study were collected at the latest follow up . The study and data collection protocols have been described in detail in [10] . The YFS study protocols have been approved by local ethics committees . The Northern Finland Birth Cohort 1966 ( NFBC66 ) has been described in detail previously [11] . The original study design focused to study factors affecting pre-term birth , low birth weight , and subsequent morbidity and mortality . Mothers living in the two northern-most provinces of Finland were invited to participate if they had expected delivery dates during 1966 . Individuals still living in the Helsinki area or Northern Finland ( N = 4 , 703 ) were asked to participate in a detailed biological and medical examination as well as a questionnaire at the age of 31 years . The NFBC66 study protocols have been approved by local ethics committees . The subjects used in the adipose tissue eQTL analysis were obtained from the EUFAM study ( European Multicenter Study of Familial Dyslipidemias ) database [30] including a Finnish cohort with familial low HDL-C phenotype . The Ethical Committee of the Department of Medicine , Helsinki University Central Hospital approved the EUFAM study . Top and bottom 10th percentiles of HDL-C concentrations ( Finnish population age and sex specific percentiles ) were used to define the high and low HDL-C groups , respectively , and subject who were not matched for BMI were removed . Subcutaneous adipose tissue biopsies were obtained from 54 individuals . Out of these , 35 individuals had low HDL-C and 19 individuals high HDL-C . Individuals in both low and high HDL-C groups were matched by age and gender . Fat biopsies were collected , RNA extracted and quantified as previously described [53] . RNA labeling , array processing and scanning were done according to the standard protocol by Affymetrix using HG-U133 ( Plus 2 . 0 ) arrays . Pre-processing of the expression data was done using GC-RMA normalization . Genotyping was performed using the HumanCNV370v1_C platform at the Broad Institute . SNPs with genotype rate <90% were excluded from the analyses and samples were removed if fewer than 95% of SNPs could be genotyped in them . In the Tampere vascular study ( TVS ) , vascular samples were collected from patients undergoing peripheral vascular surgery due to symptomatic atherosclerosis ( cerebrovascular disease due to carotid stenosis , peripheral arterial disease ) . All of these patients had a polyvascular disease which had affected at least two different vascular beds . Control samples were taken from left internal thoracic arteries ( LITA ) during coronary artery by-pass surgery ( n = 25 ) . Atherosclerotic plaques were collected by endarterectomy from the following arterial sites: femoral artery ( n = 24 ) carotid artery ( n = 29 ) and abdominal aorta ( n = 15 ) all together from a total of 93 patients . The vascular samples were classified according to the American Heart Association classification ( AHA ) [54] . The carotid and femoral artery samples were of type V or VI , aortic samples were type VI and all control vessels were macroscopically and microscopically healthy . The samples were taken from patients subjected to open vascular surgical procedures at the Division of Vascular Surgery , Tampere University Hospital . The study was approved by the Ethics Committee of Tampere University Hospital . All patients gave informed consent . The HLC and BxH-ApoE data was obtained from the Sage BioNetworks repository . A detailed description of the HLC data can be found here [31] , [32] . Detailed information on mouse experiments and sample handling can be obtained here [34] , [35] . An outlier with extreme lesion area ( Z-score = 4 . 166 , P = 1 . 5×10−5 ) was removed from analysis . Inclusion of the outlier did not affect significance . The samples from the NFBC66 , YFS and DILGOM cohorts were analyzed using the same high-throughput serum NMR metabolomics platform [17] providing information on lipoprotein subclass distribution and lipoprotein particle concentrations , low-molecular-weight metabolites such as amino acids , 3-hydroxybutyrate , and creatinine , and detailed molecular information on serum lipids including free and esterified cholesterol , sphingomyelin , saturation and ω-3 fatty acids . Further details of the NMR spectroscopy , data analyses as well as the full metabolite identifications have been described previously [17] , [29] . Individuals known to be on lipid-lowering therapy or pregnant were excluded from analysis . To calculate residuals for all metabolites , each study included the following as covariates: gender , age ( only YFS , the NFBC66 is a birth cohort ) , and loadings of the first 10 principal components from genetic data to correct for cryptic population stratification . Residuals were normalized using an inverse normal transformation to have a mean of zero and a standard deviation of 1 . In combining the metabolomic data for the YFS and NFBC66 , residuals for all metabolites also included the cohort as a covariate . Processing of metabolites from the DILGOM cohort has been described previously [29] . The YFS and NFBC66 cohorts studied were genotyped using standard protocols on the Illumina 670 BeadArray and Illumina 370CNVduo ( Illumina , Inc . San Diego , CA , USA ) respectively . Prior to imputation , stringent quality filtering was employed for each cohort . Quality control was performed independently for each study prior to imputation . Low quality SNPs ( >5% missingness ) and poor DNA samples ( >5% individual missingness ) were removed . In addition , individuals with high genomic heterozygosity ( indicative of sample contamination ) , gender discrepancies or closely related individuals were removed from the data . Genotype imputation was performed using the MACH algorithm [14] and the CEPH reference panel from HapMapII [15] . After filtering , sample numbers were 1 , 905 and 4 , 703 for the YFS and NFBC66 cohorts respectively . After imputation , SNPs were filtered within each cohort via the estimated squared correlation between imputed and true genotypes ( Rsq<0 . 30 ) , estimated minor allele frequency ( MAF<0 . 01 ) , and Hardy-Weinberg equilibrium exact test ( P<1 . 0×10−6 ) . After SNP filtering , 2 , 406 , 682 and 2 , 360 , 512 SNPs in the YFS and NFBC66 cohorts respectively were taken forward for further genome-wide analyses . The intensity cluster plots for the top , directly-genotyped SNPs were visually inspected for failures in genotype assay and calling . In order to define matrices of related endogenous variates , groupings of metabolites must be defined . Normalized metabolite measurements across the YFS and NFBC66 cohorts were pooled , and the metabolite-metabolite Pearson correlation matrix was hierarchically clustered . From the resulting dendrogram , metabolite cluster detection was done using a dynamic tree cutting algorithm [18] with a minimum cluster membership of one metabolite . We selected the dynamic tree cutting algorithm because it has been shown to outperform other popular methods in simulations as well as give biologically relevant results on real data [29] , [55] , [56] . In order to maximize power to detect associations , Ferreira and Purcell showed that , within a phenotype set , one should maximize both the number of phenotypes and the level of correlation between phenotypes [4] , however in practice these two parameters are inversely related . That is , given a of phenotype measures and individuals , increasing the number of phenotype clusters leads to increasing correlation within clusters and vice versa . For the dynamic tree cut algorithm , we investigated the sensitivity of cluster splitting using the deepSplit parameter . Lower integers values of deepSplit correspond to lower sensitivity for cluster splitting and thus fewer clusters . Both high and low sensitivity ( deepSplit = 4 and deepSplit = 0 , respectively ) for cluster splitting were explored using the YFS discovery cohort . The high setting assigned 11 metabolic networks ( Figure S1 ) whereas the low setting assigned 5 metabolic networks ( Figure S7 ) . Both clusterings were empirically assessed using the multivariate test above , and both settings detected the same number of loci at genome-wide significance . This was consistent with the inverse relationship between intra-cluster correlation and number of metabolites per cluster . Given no difference in locus detection , we considered the biological interpretation of the clusters . We noted that the low setting could not differentiate TG-rich VLDL particles nor lipid poly-unsaturation and conflated various energy metabolites with small HDL metabolism . Since it presented more straight-forward biological interpretation , we proceeded to downstream analysis with the high sensitivity , corresponding to 11 metabolic networks . Association testing of SNPs and metabolites was done using two strategies: univariate linear regression and multivariate Canonical Correlation Analysis ( CCA ) . For the former , we used the standard framework where Yi is the normalized metabolite measure for individual i , Xi is the genotype of the individual at a given SNP ( encoded as 0 , 1 , or 2 for the number of minor alleles ) , and εi is a normally distributed random variable with mean equal to zero and constant variance . To implement linear regression , we used the PLINK analysis software [57] . The reported P values assume a NULL hypothesis of no association , b = 0 . When testing hypotheses that include multiple endogenous variables , the relationships among the endogenous variables must be taken into account in addition to those between the endogenous and exogenous variables . Given these two sets of variables , the aim is to simultaneously find the best predictor of the linear functions of one set as well as the linear function of the other set it best predicts . This yields a pair of variates which are referred to as the first pair of canonical variates . Using the residuals of these linear functions , the process can be repeated to obtain the second pair of canonical variates , and so on . The full sequence of these pairs of variates and their correlations then fully describe the invariant relationships between the endogenous variable set and the exogenous variable set [58] . For multivariate testing , we use the CCA framework implemented in the PLINK . multivariate analysis tool [4] and the R statistical programming language . In this case , the exogenous set consists of only one variable , the SNP , and consequently only one pair of canonical variates is calculated . Wilks' lambda ( λ ) is a multivariate analogy of the F-test in one-way analysis of variance . In a genetic setting , λ is a statistic which tests for differences between the means of the three genotype groups ( AA , AB , and BB ) on a combination of endogenous variates ( a network of metabolite variables ) . In this case , λ = 1−ρ2 where ρ is the canonical correlation coefficient between the SNP and the network of metabolite variables . The calculation of a P-value arises from a transformation of Wilks' lambda into a statistic which is approximately F distributed . Genomic inflation of test statistics can be an indicator of subtle biases in the data and testing ( e . g . cryptic population structure ) . To assess genomic inflation , we compared our observed distribution of −log10 ( P ) values to that expected in the absence of association . A linear model was then fitted to the lowest 90% of the distribution and genomic inflation was taken as the slope of the fitted line . Table S1 gives genomic inflation values for multivariate testing of YFS , NFBC66 and meta-analysis across all metabolic networks . A locus was defined as a 200 kb genomic region centered on the top significantly associated SNP . To determine the independence of the locus-network signals , conditional multivariate association analysis was used for signals either at 4q13 . 3 , close to LIPC ( i . e . AQP9 variants ) , or close to FADS1/2/3 ( i . e . CD5/CD6 and INCENP/FTH1/BEST1 variants ) . For the top SNP ( s ) at a proximal locus ( Table S2 ) , each metabolite in a network was regressed onto the proximal SNP ( s ) and the resulting residuals were used as endogenous variables in the multivariate test of the target locus . An attenuated signal indicates non-independence , e . g . the SNPs tag the same causal variant . As a result , two loci ( CD5/CD6 and INCENP/FTH1/BEST1 ) were largely attenuated and not regarded as independent from the FADS1/2/3 locus . Due to the different number of statistical tests , genome-wide significance differs between multivariate and univariate testing . Here , as the basis of genome-wide significance , we use the common threshold of 5 . 0×10−8 , derived from the number of independent common haploblocks in genomes of European descent [59] . Univariate testing of all 130 metabolites , implies a Bonferroni corrected significance level of 5 . 0×10−8/130 = 3 . 8×10−10 . Multivariate testing of the metabolic networks we identify here ( N = 11 ) gives a Bonferroni corrected significance level of 5 . 0×10−8/11 = 4 . 5×10−9 . Analysis of the DILGOM cohort considered the bead-weighted and quantile normalized gene expression data from the Illumina HT-12 expression array as described previously [55] and metabolomic measures also described previously [29] . Only those metabolites which were part of the original metabolic network associated with a particular locus were considered . For example , because genetic variation at SERPINA1 was associated with metabolic networks 1 and 2 , expression of SERPINA1 was only tested for association with metabolites from networks 1 and 2 . Since loci were associated with different metabolic networks , different numbers of tests were performed for each candidate gene . We therefore implemented appropriate multiple testing thresholds for each gene where significance was set at P< ( 0 . 05/total_number_metabolites_in_tested_networks ) . For the EUFAM study , statistical eQTL analyses were performed using a linear regression model adjusting for the gender , BMI , and low/high HDL-C affection status . Probe intensities were treated as dependent and genotypes as independent variables . The comparison of gene-expression between the low and high HDL-C groups was performed using a linear regression model . Fold change for each probe was calculated by dividing the mean probe intensity in the low HDL-C group by the mean probe intensity in the high HDL-C group . For the TVS study , all vascular specimens were immediately frozen and RNA was extracted as previously described [60] . RNA was reverse transcribed into cRNA , biotin-UTP labelled using the Illumina TotalPrep RNA Amplification Kit ( Ambion ) and cRNA hybridized to the Illumina HumanHT-12 v3 Expression BeadChip . BeadChips were scanned with the Illumina iScan system . Data processing was conducted using R language and appropriate Bioconductor modules . Robust multi-array averaging ( RMA ) [61] was used to correct negative intensity values after background subtraction . Between arrays normalization was done using robust spline normalization ( RSN ) [61] . Quality control was performed using sample clustering and multi-dimensional scaling . Seven outliers were removed due to low expression profiles , 4 from carotid artery group and 3 from LITA group . Fold changes ( FCs ) for differentially expressed genes were calculated from log2-transformed median expression values between case ( carotid , abdominal , femoral ) and control group ( LITA ) , and the significance of the differences were evaluated with non-parametric Mann-Whitney U test due to non-normal distribution of expression values and relatively small sample sizes of TVS . If there were more than one probe presenting a gene in the expression chip , the probe with highest median expression value was selected for FC calculation .
|
In this study , we aim to identify novel genetic variants for metabolism , characterize their effects on nearby genes , and show that the nearby genes are associated with metabolism and atherosclerosis . To discover new genetic variants , we use an alternative approach to traditional genome-wide association studies: we leverage the information in phenotype covariance to increase our statistical power . We identify variants at seven novel loci and then show that our top signals drive expression of nearby genes AQP9 and SERPINA1 in multiple tissues . We demonstrate that AQP9 and SERPINA1 gene expression , in turn , is associated with metabolite levels . Finally , we show that the genes are associated with atherosclerosis using mouse atherosclerotic lesion size ( AQP9 ) as well as tissue from healthy human arteries and atherosclerotic plaques ( AQP9 and SERPINA1 ) . This study illustrates that multivariate analysis of correlated metabolites can boost power for gene discovery substantially . Further functional work will need to be performed to elucidate the biological role of SERPINA1 and AQP9 in atherosclerosis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genome-wide",
"association",
"studies",
"systems",
"biology",
"medicine",
"gene",
"expression",
"genetics",
"atherosclerosis",
"biology",
"genetics",
"and",
"genomics",
"cardiovascular"
] |
2012
|
Novel Loci for Metabolic Networks and Multi-Tissue Expression Studies Reveal Genes for Atherosclerosis
|
Numerous genetic and epigenetic alterations render cancer cells selectively dependent on specific genes and regulatory pathways , and represent potential vulnerabilities that can be therapeutically exploited . Here we describe an RNA interference ( RNAi ) –based synthetic interaction screen to identify genes preferentially required for proliferation of p53-deficient ( p53− ) human cancer cells . We find that compared to p53-competent ( p53+ ) human cancer cell lines , diverse p53− human cancer cell lines are preferentially sensitive to loss of the transcription factor ETV1 and the DNA damage kinase ATR . In p53− cells , RNAi–mediated knockdown of ETV1 or ATR results in decreased expression of the telomerase catalytic subunit TERT leading to growth arrest , which can be reversed by ectopic TERT expression . Chromatin immunoprecipitation analysis reveals that ETV1 binds to a region downstream of the TERT transcriptional start-site in p53− but not p53+ cells . We find that the role of ATR is to phosphorylate and thereby stabilize ETV1 . Our collective results identify a regulatory pathway involving ETV1 , ATR , and TERT that is preferentially important for proliferation of diverse p53− cancer cells .
The p53 tumor suppressor ( also called TP53; NP_000537 . 3 ) plays a pivotal role in regulating multiple cellular processes including cell cycle arrest , apoptosis , cell metabolism and senescence ( reviewed in [1] ) . Activated p53 can either induce cell cycle arrest and inhibit cell growth or promote cell apoptosis depending on the type of stress and the cellular context . Mutations that perturb p53 function , typically in its DNA-binding domain , or disruptions of the p53 upstream or downstream regulatory networks , have been found in more than half of all cancer cases and are present in cancer-prone families with Li-Fraumeni syndrome ( OMIM#151623 ) ( reviewed in [2] ) . In addition , loss of p53 function is often associated with increased resistance to chemotherapy and/or poor survival ( see , for example , [3]–[5] ) . For these reasons , the selective molecular targeting of p53-deficient ( p53− ) tumors has remained one of the most important goals and challenges of molecular oncology ( reviewed in [6] ) . One strategy for treating p53− tumors is to reestablish the growth-inhibitory functions of p53 . The feasibility of this approach is supported by animal studies demonstrating that reactivation of wild type p53 leads to tumor regression [7]–[9] . Several experimental strategies have been used to restore p53 activity . For example , gene therapy involving viral vectors has been used to reintroduce p53 into p53− tumor cells [10] . Alternatively , for cancers that retain a wild type p53 gene , small molecule drugs have been identified that stabilize and activate p53 by interfering with its negative regulator MDM2 ( NP_002383 . 2 ) ( reviewed in [11] ) . Restoration of p53 function in cancers expressing only mutant p53 is even more challenging . However , small molecules that refold mutant p53 proteins , and thus reactivate p53 function , have been described ( reviewed in [12] ) . Some of these approaches have progressed to clinical trials but to date none have been found to have clearly demonstrable clinical benefit [13] . An alternative approach to restoration of p53 function would be to target proteins that are preferentially required for proliferation or survival of p53− cells . Such targets can , in principle , be identified by synthetic lethal interaction screens , an idea first proposed by Hartwell et al . based upon studies in the budding yeast Saccharomyces cerevisiae [14] . The validity of this approach is supported by the realization that cancer cells are highly dependent upon or “addicted” to specific genes and regulatory pathways , confirming the existence of cancer cell-selective synthetic interaction targets ( reviewed in [15] , [16] ) . In addition , an important proof-of-principle is the demonstration that small molecule inhibitors of poly ( ADP-ribose ) -polymerase ( NP_001609 . 2 ) are synthetically lethal with loss-of-function mutations in BRCA1 ( NP_009225 . 1 ) , BRCA2 ( NP_000050 . 2 ) as well as other components of the homologous recombination pathway [17]–[19] . Here we carry out an RNA interference ( RNAi ) -based synthetic interaction screen to identify a regulatory pathway that is preferentially required for proliferation of p53− cancer cells .
To identify genes that are preferentially required for the viability and proliferation of p53− cancer cells , we designed a synthetic interaction screen , which is summarized in Figure 1A and briefly described below . The primary screen was carried out using a well-characterized isogenic pair of human HCT116 colorectal cancer cell lines , one harboring wild type p53 ( p53+ HCT116 ) and the other bearing a homozygous p53 deletion ( p53− HCT116 ) [20] . For these and all other cell lines used in this study , the presence or absence of functional p53 was confirmed by monitoring expression of the p53 target gene , p21 ( also called CDKN1A; NP_510867 . 1 ) ( Figure S1 ) . A human shRNA library comprising ∼60 , 000 shRNAs directed against ∼27 , 000 genes [21] was packaged into lentivirus particles , pooled and used to infect in parallel the two HCT116 cell lines . Ten days later , genomic DNA from both cell lines was isolated , and shRNAs were PCR amplified and subjected to massively parallel sequencing; as a reference , the starting shRNA population in both cell lines ( taken 40 hours post-infection ) was also analyzed . Statistical analysis of the four shRNA populations identified shRNAs targeting 103 genes ( Table S1 ) whose abundance was significantly decreased in p53− HCT116 cells ( ≥4-fold ) but not in p53+ HCT116 cells ( ≤2-fold ) at 10 days post-infection relative to the earlier time point ( Figure 1B ) . Such shRNAs are presumably synthetic with the p53 deletion , thus rendering p53− cells harboring these shRNAs inviable or growth impaired , and leading to their relative under-representation in the p53− HCT116 population . To validate candidates isolated from the primary screen , each shRNA was analyzed in an independent colony formation assay . p53+ and p53− HCT116 cells were infected with a lentivirus expressing a single candidate shRNA , and 10 days later cells were puromycin selected , re-plated at low density , and monitored for colony formation . This secondary screen revealed 11 genes that , when knocked down , substantially decreased colony formation in p53− HCT116 cells compared to p53+ HCT116 cells ( Figure 1C ) . ShRNAs targeting these 11 genes also preferentially decreased the ability of p53− HCT116 cells to proliferate in culture ( Figure 1D and summarized in Table S2 ) . Quantitative RT-PCR ( qRT-PCR ) confirmed in all cases that expression of the target gene was decreased in the knockdown cell line ( Figure S2A ) . To rule out the possibility that growth inhibition was due to an off-target effect of the shRNAs , for each of the 11 genes we derived a short interfering RNA ( siRNA ) whose sequence was unrelated to the original shRNA used in the experiments described above . Figure 1E shows that each siRNA also preferentially decreased proliferation of p53− HCT116 cells compared to p53+ HCT116 cells . In all cases , qRT-PCR analysis confirmed that the siRNA decreased expression of the target gene ( Figure S2B ) . p53− tumors from both the same and different types of cancers vary substantially with regard to additional genetic and epigenetic aberrations [22] . We were therefore interested in determining whether the 11 genes we identified were also preferentially required for the growth of other p53− cancer cells . To address this issue , we first analyzed an isogenic pair of human RKO colorectal cancer cell lines , one harboring wild type p53 ( p53+ RKO cells ) and the other bearing a homozygous p53 deletion ( p53− RKO cells ) ( see Figure S1 ) . ShRNAs to the 11 genes were introduced into the isogenic pair of RKO cell lines and proliferation was monitored . The results of Figure 2A indicate that five genes ( ATR [NP_001175 . 2] , ETV1 [NP_001156619 . 1] , GFPT2 [NP_005101 . 1] , NT5C3 [NP_001002009 . 1] and UMPS [NP_000364 . 1] ) were preferentially required for growth of p53− RKO cells compared to p53+ RKO cells . By contrast , knockdown of the other six genes did not substantially inhibit growth of either p53− or p53+ RKO cells and were thus not further analyzed . We next examined these five candidates in an unrelated isogenic pair of A549 human lung cancer cell lines . In this case , the parental p53+ A549 cells were rendered p53− by stable expression of a p53 dominant-negative mutant [23] ( see Figure S1 ) . The results of Figure 2B show that siRNAs against the five candidate genes ( ATR , ETV1 , GFPT2 , NT5C3 and UMPS ) preferentially inhibited growth of the p53− A549 cell line . Finally , we analyzed the five candidate genes in a panel of four human lung cancer cell lines , two of which expressed wild type p53 ( A549 and NCI-H460 ) and two of which were compromised for p53 function ( NCI-H1299 , which lacks p53 , and NCI-H522 , which expresses a p53 mutant ) ( see Figure S1 ) . Of the five candidate genes , knockdown of two , ATR and ETV1 , were the most consistent in preferentially inhibiting proliferation of p53− cell lines ( Figure 2C ) and were selected for further analysis . ATR encodes a checkpoint kinase involved in the DNA damage response [24] , and ETV1 encodes a member of the ETS family of transcription factors [25] . We also tested whether knockdown of ATR and ETV1 would preferentially inhibit growth of p53− HCT116 tumors in a mouse xenograft model . p53+ or p53− HCT116 cells expressing an shRNA against ATR or ETV1 , or a control non-silencing shRNA , were injected subcutaneously into opposite flanks of the same nude mouse , and tumor growth was monitored after four weeks . As expected , the control p53− HCT116 cells formed larger tumors than their p53+ counterparts ( Figure 2D ) . Notably , knockdown of ATR or ETV1 markedly inhibited growth of p53− HCT116 tumors but did not have a significant effect on growth of p53+ HCT116 tumors . We next sought to investigate the basis by which ETV1 and ATR were preferentially required for growth of p53− cells . A previous study has shown that ETV1 is a transcriptional activator of TERT ( NP_001180305 . 1 ) [26] , which encodes the catalytic subunit of telomerase and has a well-established role in the maintenance of cellular proliferation [27] . Therefore , in the first set of experiments we analyzed the effect of depleting ETV1 as well as ATR on TERT levels . Significantly , RNAi-mediated knockdown of ETV1 or ATR resulted in a substantial decrease in TERT protein ( Figure 3A ) and mRNA ( Figure 3B ) levels in p53− HCT116 cells but unexpectedly had only a modest effect on TERT levels in p53+ HCT116 cells . The effect of knockdown of both ETV1 and ATR in p53− HCT116 cells on cellular proliferation and TERT levels was similar to that observed with single knockdowns ( Figure S3A ) . Pharmacological inactivation of ATR using two different chemical inhibitors , CGK773 [28] and ETP46464 [29] , also resulted in decreased TERT levels in p53− but not p53+ HCT116 cells ( Figure 3C ) . Inhibition of ATR was confirmed by monitoring phosphorylation of its target substrate , CHK1 ( also known as CHEK1; NP_001107593 . 1 ) ( Figure S4 ) . We also monitored senescence induction and performed cell cycle analysis in cells depleted of ETV1 or ATR . Figure 4A shows that in both p53+ and p53− HCT116 cells , RNAi-mediated knockdown of TERT substantially increased the number of cells that stained positively for senescence-associated β-galactosidase activity , indicative of senescence induction ( see also Figure S5A ) . The level of senescence was higher in p53+ HCT116 TERT knockdown cells than in p53− HCT116 TERT knockdown cells , as expected , because p53-directed pathways contribute to senescence [1] . Significantly , Figure 4B shows that RNAi-mediated knockdown of ETV1 or ATR also induced senescence ( see also Figure S5B ) . However , following knockdown of ETV1 or ATR , the induction of senescence was much greater in p53− HCT116 cells compared to p53+ HCT116 cells ( Figure 4B and Figure S5C ) , consistent with the difference in TERT levels ( see Figure 3A ) . In addition , knockdown of TERT increased the percentage of p53− HCT116 cells but not p53+ HCT116 cells in G2/M ( Figure 4C and Figure S6A ) . Notably , a similar preferential increase in the percentage of p53− HCT116 cells in G2/M occurred following knockdown of ETV1 or ATR ( Figure 4D and Figure S6B , S6C ) . To determine whether decreased TERT levels were responsible for the preferential growth defect in p53− HCT116 cells depleted of ETV1 or ATR , we performed a “rescue” experiment . Proliferation was measured both by an Alamar Blue assay ( Figure 4E ) and by quantifying cell number ( Figure S7A ) following knockdown of ETV1 or ATR in p53+ and p53− HCT116 cell lines stably expressing either TERT or , as a control , green fluorescence protein ( GFP ) ( Figure S7B ) . The results of Figure 4E and Figure S7A show that ectopic expression of TERT counteracted the large decrease in proliferation that occurred following knockdown of ETV1 or ATR in p53− HCT116 cells , as well as the modest proliferative decrease following ETV1 knockdown in p53+ HCT116 cells . Thus , the reduced TERT levels following depletion of ETV1 or ATR can largely explain the decreased proliferation of p53− HCT116 cells . Consistent with this conclusion , TERT knockdown had a similar effect on proliferation of p53− cell lines to that observed following knockdown of ETV1 or ATR ( see Figure S3 ) . Moreover , knockdown of both ETV1 and TERT , or ATR and TERT , decreased proliferation of p53− cell lines similarly to that observed with single knockdowns ( Figure S3 ) . As described above , ETV1 has been previously shown to transcriptionally stimulate TERT expression [26] . However , the basis by which ATR promotes TERT expression is unknown . The similar results obtained with ATR and ETV1 in the TERT expression experiments of Figure 3 ( and Figure S3 ) raised the possibility that the two proteins act in a common pathway . To address this possibility , we first asked whether ATR regulates ETV1 levels . The immunoblot results of Figure 5A show that in both p53+ and p53− HCT116 cells knockdown of ATR resulted in reduced ETV1 protein levels . The qRT-PCR results of Figure 5B showed that ATR depletion did not lead to reduced ETV1 mRNA levels , indicating that the ATR-mediated regulation of ETV1 occurs post-transcriptionally . Addition of an ATR chemical inhibitor also led to reduced ETV1 protein levels in both p53+ and p53− HCT116 cells ( Figure 5C ) . Following inhibition of ATR activity , DNA damage did not result in ETV1 stabilization ( Figure S8 ) . To test the generality of these results , we analyzed the effect of ATR pharmacological inhibition in several of the p53+ and p53− cell lines described above . Figure 5D shows that ATR inhibition reduced TERT protein levels in p53− RKO , NCI-H522 and NCI-H1299 cells but not in p53+ RKO , A549 and NCI-H460 cells Moreover , addition of an ATR inhibitor reduced ETV1 levels to varying extents in all cell lines . TERT protein levels were also reduced following knockdown of ATR or ETV1 in NCI-H522 cells and two additional human cancer cell lines that express mutant p53 ( Figure S9A ) , as well as in HeLa cells , which lack p53 activity due to expression of the human papilloma virus E6 protein ( Figure S9B ) . Thus , the results in these other p53+ and p53− cell lines are similar to those obtained in the isogenic pair of HCT116 cell lines used throughout this study . Because ATR is a protein kinase , a likely mechanism for the ability of ATR to post-transcriptionally regulate ETV1 is through direct interaction and phosphorylation . Consistent with this possibility , ETV1 contains five potential ATR phosphorylation sites ( Figure 6A ) . To test this idea , we ectopically expressed a FLAG-tagged ETV1 derivative ( Figure S10 ) in p53+ and p53− HCT116 cells , and analyzed interaction between FLAG-ETV1 and ATR in a co-immunoprecipitation assay . The results of Figure 6B show that in both p53+ and p53− HCT116 cells , FLAG-ETV1 could be detected in the ATR immunoprecipitate ( left ) and , conversely , ATR could be detected in the FLAG immunoprecipitate ( right ) , indicating ATR and ETV1 physically associate . To determine whether ETV1 was an ATR substrate , we immunoprecipitated FLAG-ETV1 from transfected p53+ and p53− HCT116 cell lysates and analyzed the immunoprecipitate by immunoblotting with an antibody that recognizes a phosphorylated serine followed by a glutamine [30] , the product of ATR or ATM phosphorylation [31] , [32] . The results of Figure 6C show that the immunoprecipitated FLAG-tagged ETV1 could be detected by the ATM/ATR phospho-specific antibody , suggestive of phosphorylation by ATR . Moreover , following treatment of cells with an ATR inhibitor , the immunoprecipitated FLAG-tagged ETV1 was no longer detected by the ATM/ATR phospho-specific antibody ( Figure S11 ) . To confirm that ATR phosphorylates ETV1 , we performed in vitro kinase experiments . We first tested whether ATR , in the presence of its positive effector ATRIP ( NP_569055 . 1 ) [33] , [34] , could phosphorylate a glutathione-S-transferase ( GST ) -ETV1 ( amino acids 1–290 ) fusion-protein that contained all five potential ATR phosphorylation sites . The results of Figure 6D show that ATR phosphorylated the GST-ETV1 fusion-protein but , as expected , not a control GST protein . To confirm and extend this result , we constructed and analyzed a series of GST-ETV1 fusion-proteins each containing a single potential ATR phosphorylation site . The results of Figure 6E show that only one of the five potential ATR phosphorylation sites ( SQ2 ) was a substrate for ATR . Collectively , the results described above indicate that ATR phosphorylates ETV1 and stabilizes it from proteolytic degradation . As discussed above , previous studies have shown that ETV1 is a transcriptional activator of TERT [26] . Therefore , we thought the most likely mechanism by which ETV1 promotes proliferation in p53− HCT116 cells is through direct binding to the TERT promoter and stimulation of TERT transcription . To test this possibility , we performed chromatin-immunoprecipitation ( ChIP ) experiments . The ChIP experiments of Figure 7A ( left panel ) show that in p53− HCT116 cells , ETV1 was bound to a region within intron 1 , which has been previously reported to contain multiple ETV1 binding sites and is required for complete TERT transcriptional activity [26] . Remarkably , in p53+ HCT116 cells , whose proliferation is not dependent upon ETV1 , there was no detectable binding of ETV1 to the same region of the TERT promoter . Notably , ectopic expression of wild type p53 in p53− HCT116 cells resulted in substantially decreased binding of ETV1 to the TERT promoter ( Figure 7B , left ) . Conversely , ectopic expression of a p53 dominant-negative mutant in p53+ HCT116 cells resulted in substantially increased binding of ETV1 to the TERT promoter ( Figure 7B , right ) . In p53− HCT116 cells , binding of ETV1 to the TERT promoter was lost following pharmacological inhibition of ATR ( Figure 7A and Figure S12A ) , which as shown above results in decreased ETV1 levels ( see Figure 5C ) . Conversely , binding of ETV1 to the TERT promoter modestly increased following irradiation with ultraviolet light , which increases ATR activity ( Figure S12B ) . ChIP experiments monitoring ATR occupancy revealed that ATR was bound to the same region of the TERT promoter as ETV1 ( Figure 7C ) . Thus , in p53− HCT116 cells , ETV1 and ATR are both bound to the TERT promoter , which is consistent with our finding that the two proteins are physically associated ( Figure 6B ) . In conjunction with a previous study [26] , the results presented above suggested that ETV1 is directly responsible for stimulating TERT expression and that ATR functions by phosphorylating and thereby stabilizing ETV1 . A prediction of this model is that ectopic expression of ETV1 would bypass the requirement of ATR for proliferation of p53− HCT116 cells . The rescue experiment of Figure 7D shows that the decreased proliferation of p53− HCT116 cells following knockdown of ATR was counteracted by ectopic expression of ETV1 ( Figure S13 ) . Following knockdown of TERT , ectopic expression of ETV1 could no longer rescue proliferation of p53− HCT116 cells depleted of ATR ( Figure S14A ) . In these experiments , ectopic expression of ETV1 had no effect on γ-H2AX foci formation , a marker of DNA damage [35] ( Figure S14B ) . These results suggest that the growth arrest observed following loss of ATR is primarily due to decreased ETV1 levels .
In this report we have performed a large-scale shRNA screen to identify a regulatory pathway involving ETV1 , ATR and TERT that is preferentially required for proliferation of diverse p53− cancer cells . We found that in p53− cells , TERT transcription is highly dependent upon ETV1 , which functions as a direct transcriptional activator by binding to the TERT promoter downstream of the transcription start-site . In p53+ cells , ETV1 , although present at comparable levels , is not required for TERT transcription and surprisingly is not bound to the same region of the TERT promoter . Notably , ectopic TERT expression restored normal proliferation in p53− cells depleted of ETV1 or ATR ( Figure 4E and Figure S7A ) , indicating that the promotion of TERT expression is an important , but not necessarily the only , mechanism by which ETV1 and ATR maintain proliferation of p53− cells . Consistent with our results , a previous study reporting a requirement for ETV1 in TERT transcription [26] was primarily based upon experiments performed in 293T cells , which lack p53 activity due to expression of SV40 large T antigen . The results described above suggest that p53+ cells express a transcription factor that functionally substitutes for ETV1 , and that one or more proteins associated with the TERT promoter in p53+ cells prevent binding of ETV1 . Several transcription factors , including SP1 ( NP_612482 . 2 ) , E2F1 ( NP_005216 . 1 ) and MYC ( NP_002458 . 2 ) , have been previously shown to be associated with the human TERT promoter ( reviewed in [36] ) . To ask whether these factors , or p53 itself , might contribute to the differential regulation of TERT we performed ChIP experiments in p53+ and p53− HCT116 cells . Consistent with previous studies , we found that E2F1 and MYC were associated with the TERT promoter; binding of E2F1 was modestly increased in p53− HCT116 cells ( Figure S15A ) , whereas for MYC there was no difference in p53+ and p53− HCT116 cells ( Figure S15B ) . In p53+ HCT116 cells there was increased binding of SP1 ( Figure S15C ) and , most notably , there was substantial binding of p53 to the TERT promoter ( Figure S15D ) . Interestingly , a number of previous studies have reported physical and functional interactions between SP1 and p53 ( see , for example , [37]–[41] ) . Our ChIP results reveal substantial differences between the composition of proteins associated with the TERT promoter in p53+ and p53− HCT116 cells , which may be related to the differential requirement for ETV1 . Interestingly , in contrast to human cancer cell lines , we found that ATR was not required for TERT expression in experimentally derived p53− MCF10A cells , an immortalized but non-transformed human cell line ( Figure S16A ) . In addition , ATR was not required for TERT expression in p53− mouse embryo fibroblasts ( Figure S16B ) , consistent with the lack of conservation between the mouse and human TERT promoter ( data not shown ) . Thus , the requirement of ATR and ETV1 for TERT expression may be specific to human p53− cancer cell lines . Several previous studies have reported results that are consistent with the synthetic interaction between p53 and ATR we have described here . For example , p53− cells have been found to be particularly sensitive to pharmacological inhibition of ATR ( see , for example , [42]–[44] ) . In addition , mice expressing a hypomorphic allele of ATR have an aging phenotype that is exacerbated in the absence of p53 [45] . Significantly , mouse embryo fibroblasts containing this hypomorphic ATR allele have an elongated G2 phase following loss of p53 , consistent with our cell cycle results ( Figure 4D and Figure S6B , S6C ) . However , a preferential role for ETV1 in p53− cells and its cooperative function with ATR has not been previously described and underscores the power of unbiased , large-scale RNAi-based screens . Our screening strategy did not emphasize reaching saturation but rather sought to follow-up , by directed experiments , a limited number of candidates isolated in the primary screen . For several reasons , we believe that our screen , like other large-scale shRNA screens ( see , for example , [46] ) , did not achieve saturation . For example , a previous siRNA screen identified several factors , in particular the serine/threonine kinase receptor-associated protein UNRIP ( also called STRAP; NP_009109 . 3 ) , whose loss affected proliferation of p53− HCT116 cells more severely than p53+ HCT116 cells [47] . However , we did not isolate UNRIP in our primary screen and , conversely , ATR and ETV1 were not isolated in the previous siRNA screen , suggesting that neither screen was truly saturating . Reasons for a failure to reach saturation in this and other large-scale shRNA screens include suboptimal efficacy of some shRNAs [48] , unequal representation of shRNAs in the primary screen , and an insufficient depth of deep sequencing . Thus , it is possible that additional factors that act in the ATR-ETV1-TERT pathway , or unrelated pathways preferentially required for proliferation of p53− cells , remain to be identified . The decreased proliferation of p53− cell lines was first evident within a few days following knockdown of ETV1 , ATR or TERT . It therefore seems likely that this reduced proliferation is not a result of replicative senescence due to telomere attrition , which would require many cell divisions . Senescence occurred at much later times ( 10–14 days ) and may be a secondary effect of the proliferation block . We observed that knockdown of ETV1 , ATR or TERT resulted in an increased percentage of cells in G2/M ( Figure 4C , 4D and Figure S6 ) . Although senescent cells are generally believed to arrest in G1 , it has been found that senescent cells can also arrest in G2/M ( see , for example , [49] ) . A variety of previous studies have shown that TERT can promote proliferation by multiple mechanisms , several of which are unrelated to telomere length including inhibiting apoptosis [50] , regulating cell signaling pathways and/or stimulating expression of diverse growth-promoting genes ( see , for example , [51]–[54] ) . It seems likely that the decreased proliferation of p53− cells following depletion of ETV1 , ATR or TERT involves one of these alternative mechanisms . We have found that p53− cells depleted of ETV1 , ATR or TERT have multiple growth defects including increased levels of senescence ( Figure 4A , 4B and Figure S5 ) and an altered cell cycle ( Figure 4C , 4D and Figure S6 ) . A further understanding of how TERT promotes proliferation of p53− cells is likely to identify new factors that are potential therapeutic targets .
Animal experiments were performed in accordance with the Institutional Animal Care and Use Committee ( IACUC ) guidelines . Isogenic p53+ and p53− HCT116 and RKO cell lines [20] were provided by B . Vogelstein; A549 , NCI-H460 , NCI-H522 , NCI-H1299 and HT29 cells were obtained from the National Cancer Institute; and DLD-1 , HeLa and MCF10A cells were obtained from the American Type Culture Collection . The basis for the p53− status in each of the p53− cell lines is provided in Table S3 . p53+ and p53− mouse embryonic fibroblasts were isolated from wild type and p53−/− C57BL/6 mice . All cells were grown according to the supplier's recommendations . Stable A549 and MCF10A cell lines expressing p53-DD , which harbors a deletion of 288 amino acids ( Δ15-301; [23] ) were generated by transfection with the plasmid pBABE-hygro-p53DD ( Addgene; [55] ) or the control vector , pBABE-hygro , and selection with hygromycin ( 150–200 µg/ml ) . Stable p53+ and p53− HCT116 cell lines expressing TERT were generated by transfection with the plasmid pWZL-Blast-Flag-HA-hTERT ( Addgene; [56] ) or control plasmid pWZL-Blast-GFP ( Addgene; [57] ) , and selection with blasticidin ( 10 µg/ml ) . The ETV1 expression vector was generated by subcloning ETV1 cDNA ( Open Biosystems ) into pEF6-Blast-3xFlag to create pEF6-Blast-3xFlag-ETV1 . The pEF6-Blast-3xFlag vector was generated by cloning a BsiWI-EcoRI double-stranded oligo coding for 3xFlag-tag ( MDYKDHDGDYKDHDIDYKDDDDKEF ) in Kpn1-EcoR1-digested pEF6/V5-HIS B ( Invitrogen ) . Stable p53+ and p53− HCT116 cell lines expressing ETV1 were generated by transfection with pEF6-Blast-3xFlag-ETV1 or vector only and selection with blasticidin ( 10 µg/ml ) . The Open Biosystems GIPZ lentiviral human shRNAmir library was obtained through the University of Massachusetts Medical School RNAi Core Facility . Twelve lentiviral pools , each comprising ∼5000 shRNA clones , were generated with titers of ∼2×106 pfu/ml . These lentiviral stocks were produced following co-transfection with the packaging mix into the 293T packaging cell line . To carry out the screen , p53+ and p53− HCT116 cells were plated at 1×106 cells per 100 mm plate , transduced the next day with one shRNA pool per plate at a multiplicity of infection ( MOI ) of 1 , and grown in the absence of puromycin selection . Forty hours after transduction , 75% of cells were transduced ( as evidenced by GFP fluorescence; the marker turboGFP is present in the pGIPZ vector ) . Each plate was divided into two populations: half of the cells were pooled and genomic DNA was extracted ( referred to as “T0” ) , whereas the other half were transferred to 150 mm plates and passaged by 4-fold dilutions for 10 days , at which point the cells were pooled and the genomic DNA was extracted ( referred to as “T10” ) . To analyze the frequency of individual shRNAs in the four populations , 72 µg of genomic DNA was used as the substrate ( split into 24 tubes ) and PCR amplified ( 94°C for 1 min; 15 cycles of 94°C for 1 min , 58°C for 1 min , 72°C for 45 sec; 72°C for 10 min; and hold at 4°C ) with primers GIPZF ( 5′-GAGTTTGTTTGAATGAGGCTTCAGTAC-3′ ) and GIPZHR ( 5′-CGCGTCCTAG GTAATACGAC-3′ ) . The PCR product was gel purified , and 50 ng of DNA was used as the substrate for a second PCR amplification ( 94°C for 1 min; 15 cycles of 94°C for 1 min , 50°C for 1 min , 72°C for 45 sec; 72°C for 10 min; and hold at 4°C ) using primers Forward Acu1 primer AMN ( 5′-CAACAGAAGGCTCCTGAAGGTATATTGCTGTTGAC-3′ ) and Reverse Acu1 primer AMN ( 5′-AAATTTAAACTGAAGTACATCTGTGGCTTCACTA-3′ ) . Next , 1 µg of the PCR product was digested to completion with AcuI ( New England Biolabs ) . The digested product was then ligated to the following pre-annealed adapters: L1ShSolexA ( /5Bio/-ACACTC TTTCCCTACACGACGCTCTTCCGATCTCA ) and L1ShSolexB ( /5Phos/′-AGATCGGAAGA GCGTCGTGTAGGGAAAGAGTGT/3AmM , and L2ShSolexB ( /5Phos/-AGATCGGAAGAGC TCGTATGCCGTCTTCTGCTTG/3Bio/ ) and L2ShSolexA ( /5AmMC6/-CAAGCAGAAGACG GCATACGAGCTCTTCCGATCTAC ) . The product of the 3-way ligation was run on a 3% TAE agarose gel , visualized with ethidium bromide , purified and used as a substrate for a 15-cycle PCR reaction using Solexa-Illumina primers 1 . 1 and 2 . 1 and the cycling conditions recommended by the manufacturer . The library was analyzed using the Solexa-Illumina GA Massively Parallel Deep Sequencer . Sequence information was extracted from the image files using the Solexa-Illumina Firecrest and Bustard applications . Prior to alignment of the sequence reads , a custom Perl script was used to identify the first six bases flanking the informative sequence in 5′ and the six bases flanking the informative sequence in 3′ , starting at position 28 . The core 21 bp sequences were extracted and mapped to the human reference genome sequence ( hg18 ) using the Solexa-Illumina ELAND algorithm , allowing up to two mismatches to the reference sequence . No further analysis was performed on reads that did not contain the six bases of the 5′ sequence or the six bases of 3′ adapter sequence . Sequences mapping to the same genomic location were binned and the count for each of the mapped genomic sequences was calculated for each of the four treatments . For each of the mapped genomic sequences , the Fisher Exact Test was applied to assess whether there was a differential depletion/enrichment of the shRNA sequences between T0 and T10 for both the p53− and p53+ HCT116 cell lines . The odds ratio and its 95% confidence interval were computed for each of the mapped genomic sequences using Fisher test function in R v2 . 8 based on conditional maximum likelihood estimation . To adjust for multiplicity , B–H method [58] was used . Those shRNAs with an adjusted p-value<0 . 01 and a decrease of at least four-fold at T10 compared with T0 in p53− HCT116 cells and no more than two-fold in p53+ HCT116 ( or adjusted p-value≥0 . 01 ) were identified . The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus [59] and are accessible through GEO Series accession number GSE15967 ( http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? acc=GSE15967 ) . Lentiviral supernatants corresponding to individual shRNAs ( listed in Table S4 ) were generated in 293T cells as described above . p53+ and p53− HCT116 cells were transduced with each lentiviral preparation at an MOI of 0 . 2–0 . 4 , and grown for 10 days without puromycin selection , during which cells were passaged at a 1∶6 ratio every 4 days . Cells were then subjected to puromycin selection ( 1 . 5 µg/ml ) for 5 days . For colony formation assays , cells were split at a 1∶200 ratio and plated in 6-well plates in the presence of 1 . 5 µg/ml puromycin . After 6–7 days , cells were fixed with 4% paraformaldehyde in phosphate buffered saline ( PBS ) at 4°C overnight and then stained with 0 . 1% crystal violet in PBS to visualize the colonies . At least two independent infections were performed for each shRNA; representative images are shown . For shRNA-based experiments , cells were transduced with lentiviral supernatants at an MOI of 0 . 2–0 . 4 , and subjected to puromycin selection ( 1 . 5 µg/ml ) for 5 days . Cells were then passaged at a 1∶8 ratio every 3 days and cultured in growth medium containing puromycin . After 4 passages , cells were split at a 1∶6 ratio and seeded in a 12-well plate in RPMI medium without phenol red and supplemented with 5% fetal calf serum . After 18 h , the medium was replaced with 500 µl of medium containing 10% of an Alamar Blue solution ( Invitrogen ) . After 2 h , 100 µl of the medium was used to measure fluorescence by excitation at 530 nm and emission at 590 nm . For siRNA-based experiments , siRNA duplexes were transfected into cells using Lipofectamine RNAiMax Transfection Reagent ( Invitrogen ) according to the manufacturer's instructions . Briefly , 1 . 2 µl of Lipofectamine was complexed with the siRNA ( 40 nM final concentration ) , and the solution was diluted with 100 µl of medium and applied to 2×104 cells in a volume of 500 µl culture medium per well in 24-well plates . The medium was changed after 24 h and proliferation assessed by Alamar Blue fluorescence after an additional 72 h . Sequences of siRNAs are listed in Table S4; the control LMNA siRNA sequence was previously described [60] . All experiments were performed at least 2–3 times in either duplicate or triplicate . Four days post-siRNA-transfection , cells were trypsinized , resuspended in 0 . 5 ml growth medium , and stained with 0 . 5 ml 0 . 1% Trypan blue solution ( HyClone Trypan blue , Thermo Fisher Scientific ) . Viable cells were counted using a Countess Automated Cell Counter ( Life Technologies ) . Two independent transfections were carried out and analyzed in duplicate . 2×106 shRNA-transduced p53+ or p53− HCT116 cells were suspended in 100 µl of serum-free RPMI and injected subcutaneously into the opposite flanks of n = 9 ( for non-silencing and ATR shRNAs ) or n = 5 ( for ETV1 shRNA ) athymic Balb/c ( nu/nu ) mice ( Taconic ) . Tumor dimensions were measured every week and tumor volume was calculated using the formula π/6× ( length ) × ( width ) 2 . A Mann-Whitney test was used to determine whether knockdown of ATR or ETV1 changes the tumor volume at week 4 compared to a non-silencing shRNA . Cell extracts were prepared by lysis in modified RIPA buffer ( 0 . 05 M Tris-Cl [pH 8 . 0] , 0 . 15 M NaCl , 1% Nonidet P-40 , 0 . 5% desoxycholate , 0 . 1% SDS , 2 mM phenylmethylsulphonyl fluoride ( PMSF ) , 20 µg/ml aprotinin , 1 mM Na3VO4 and 1 mM NaF ) in the presence of a proteinase inhibitor cocktail ( Roche ) . Blots were probed with α-TERT ( Epitomics , 1531-1 ) . α-ETV1 ( Abcam , ab81086 ) , α-Flag M2 ( Sigma , F1804 ) , α-phospho-CHK1 ( Ser317 ) ( Cell Signaling Technology , 8191 ) , α-p21 ( BD Pharmingen , SX118 ) , α-tubulin ( Sigma , B5-1-2 ) or β-actin ( Sigma , AC74 ) . For ATR inhibition , cells were treated with 2–6 µM CGK733 ( Calbichem ) or 0 . 5–8 µM ETP46464 ( [29]; kindly provided by O . Fernandez Capetillo ) for 72 h prior to cell extract preparation; as a control , cells were treated with dimethyl sulfoxide ( DMSO ) . For p53 functional assays ( Figure S1 ) , cells were treated with 25 µM etoposide ( Sigma ) or 10 µg/ml 5-fluorouracil ( Sigma ) for 24 h , and cell extracts were prepared as above . For RNAi experiments , experiments were performed at least 2–3 times in either duplicate or triplicate . Total RNA was extracted using TRIzol Reagent ( Invitrogen ) and treated with Turbo DNA-free kit ( Ambion Inc . ) . The same amount of total RNA ( 3 µg ) for each sample was employed to produce templates for SYBR-green quantitative PCR analysis using SuperScript II Reverse transcriptase ( Invitrogen ) . Target genes were amplified using specific primers and expression levels were normalized to that of GAPDH . Primer sequences are listed in Table S4 . All experiments were performed at least 2–3 times in either duplicate or triplicate . Assays were performed as described previously [61] with minor modifications . Briefly , 10–14 days following RNAi-mediated knockdown , cells were washed twice with PBS , then fixed using 3 . 7% paraformaldehyde for 5 min at room temperature . After three washes with PBS , cells were incubated with fresh staining solution ( 40 mM citric acid/Na2HPO4 pH 6 . 0 , 150 mM NaCl , 2 mM MgCl2 , 5 mM potassium ferricyanide , 5 mM potassium ferrocyanide , 1 mg/ml X-Gal ) for 12–18 hr at 37°C ( no CO2 ) and covered from light . Images were captured using a Spot TE-200 digital camera ( SPOT Imaging Solutions ) . The number of blue cells in 10 fields ( each containing 100–250 cells ) was counted manually , and the percentage calculated . Two independent infections were performed for each knockdown . Cells transduced with shRNAs were harvested by trypsinization , fixed in 80% ethanol and stored at −20°C overnight . Fixed cells were stained with propidium iodide buffer containing 50 µg/ml RNase ( Sigma ) and 50 µg/ml propidium iodide ( Sigma ) in PBS . Flow cytometry was performed by the UMass Medical School Core Flow Cytometry Lab using a FACScalibur flow cytometer ( Becton Dickinson ) . Data were analyzed with FlowJo ( Tree Star ) . All experiments were performed at least 2–3 times . For Figure 6B , 5×107 p53+ or p53− HCT116 cells expressing Flag-ETV1 were rinsed twice with cold PBS , lysed in 1 ml IP lysis buffer ( 50 mM Tris-Cl pH 7 . 4 , 250 mM NaCl , 5 mM EDTA , 0 . 2%Triton X-100 , 0 . 5 mM DTT , 1× complete protease inhibitor [Roche] , and phosphatase inhibitor cocktails 2 and 3 [Sigma , p5726 and p0044] ) on ice . The lysate was cleared by centrifugation at 16 , 000 g for 30 min at 4°C . Whole cell lysate ( 2 mg per sample ) was incubated with relevant antibodies ( α-ATR [Abcam , ab2905] or control rabbit IgG [Abcam , ab37415] or α-Flag M2 [Sigma] or control mouse IgG [Santa Cruz , sc2343] ) overnight at 4°C after being precleared with 50 µl Dynabeads-protein G ( Invitrogen ) . Dynabeads Protein G ( 50 µl ) were added to each lysate-antibody complex , incubated for 2 h , spun , and washed 5 times with IP lysis buffer . Protein complexes were eluted by boiling with Laemmli buffer . For Figure 6C , immunoprecipitations were carried out as described above with α-Flag M2 ( Sigma ) , then immunoblotted with α-SQ2 ( [30]; kindly provided by S . Elledge ) , or α-ETV1 ( Abcam ) . To create GST-ETV1 ( amino acids 1–290 ) , the corresponding portion of ETV1 was PCR amplified using pEF6-Blast-3xFlag-ETV1 as a template and cloned into pGEX-4T-3 ( GE Healthcare ) . The construct was confirmed by sequencing . For the smaller GST-ETV1 fusion proteins , synthetic oligos corresponding to amino acids 9–23 ( SQ1 ) , 44–58 ( SQ2 ) , 97–111 ( SQ3 ) , 165–179 ( SQ4 ) or 240–254 ( SQ5 ) were annealed and cloned into pGEX-4T-3 . In vitro kinase assays were performed as previously described [62] except that reaction volumes were quadrupled . 32P-labeled products were visualized by autoradiography . ChIP assays were carried out as described previously [63] , [64] with the following minor modifications . Briefly , 5×107 cells were first incubated with ethylene glycol bis ( succinimidyl succinate ) ( EGS ) for 30 min and then incubated with 1% formaldehyde for 10 min at room temperature before crosslinking was quenched by addition of 0 . 125 M glycine . Cells were collected by centrifugation and lysed in lysis buffer containing 50 mM Tris–HCl pH 8 . 0 , 10 mM EDTA , 0 . 5% SDS , proteinase inhibitors ( Roche ) and phosphatase inhibitors ( Sigma ) . The cell suspension was sonicated for 15 min total time with 30 seconds ON and 30 seconds OFF using Bioruptor ( Diagenode ) . Sonicated chromatin was then incubated at 4°C overnight with 5 µg of the appropriate antibody: α-ATR ( Abcam ) , α-ETV1 ( Abcam ) , α-E2F1 ( Santa Cruz ) , α-MYC ( Cell Signaling Technology ) , α-p53 ( Santa Cruz ) , α-SP1 ( Abcam ) , and corresponding IgG control . Immunoprecipitated chromatin DNA was analyzed by real-time PCR using the following primers: TERT promoter ( −3 kb ) ( for 5′-ACGATGGAGGCAGTCAGTCT-3′; rev 5′-T CCCCACACACTTCATGCTA-3′ ) , TERT promoter ( −300 bp ) ( for 5′-GTTCCCAGGGCCTCCA CATC-3′; rev 5′-GCGGAGAGAGGTCGAATCGG-3′ ) , TERT intron 1 ( 0 . 4 kb ) ( for 5′-GAACC AGCGACATGCGGAGAGCA-3′; rev: 5′-AGCTCCTTCAGGCAGGACACCT-3′ ) . Fold enrichment was calculated by comparing the amplification threshold ( Ct ) value of a given ChIP sample with that obtained in the IgG control at the same target locus . All experiments were performed at least 2–3 times in either duplicate or triplicate . Quantification of γ-H2AX-positive cells was performed as previously described [65] with modifications . Briefly , cells were seeded onto 22-mm glass coverslips and 48 h later , the coverslips were washed in PBS , incubated in cytoskeleton buffer ( 10 mM piperazine-N , N′-bis[2-ethanesulfonic acid] [PIPES] pH 6 . 8 , 100 mM NaCl , 300 mM sucrose , 3 mM MgCl2 , 1 mM EGTA , 0 . 5% Triton X-100] for 5 min on ice . After several washes with ice-cold PBS , the cells were fixed in 4% paraformaldehyde for 20 min and permeabilized in 0 . 5% Triton X-100 solution for 15 min at room temperature . Cells were blocked with 2% BSA in PBS , incubated with primary antibody anti-phosphoH2AX ( Ser139 ) ( Millipore , JBW301 ) overnight at 4°C , washed three times with 1× PBS , and incubated with secondary antibody Cy3-conjugated sheep anti-mouse IgG ( Sigma-Aldrich ) for 1 h at room temperature . Cells were then washed , counterstained with 4′ , 6′-diamidino-2-phenylindole ( DAPI ) , and mounted in 90% glycerol and 2% 1 , 4-diaza-bicyclo- ( 2 , 2 , 2 ) -octane ( DABCO ) . Images were captured using a Zeiss AxioCam HRc camera , and 10 fields of cells were counted for each sample in duplicate .
|
The conversion of a normal cell into a cancer cell involves activating genes that promote cancer growth ( oncogenes ) and/or inactivating genes that normally act to inhibit cancer growth ( tumor suppressor genes ) . The tumor suppressor gene p53 is the most frequently mutated gene in human cancers , being inactivated in approximately half of all tumors . In addition , loss of p53 function is often associated with increased resistance to chemotherapy and/or poor survival . For these reasons , the selective destruction of p53-deficient ( p53− ) tumors has remained one of the most important goals and challenges of cancer therapy . One strategy for destroying p53− tumors is to inactivate genes that are preferentially required for the growth or survival of p53− cells . Here we carry out a large-scale genetic screen to identify a cellular pathway that is preferentially required for growth of p53− cancer cells .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"functional",
"genomics",
"rna",
"interference",
"protein",
"interactions",
"dna-binding",
"proteins",
"dna",
"transcription",
"gene",
"function",
"animal",
"models",
"mitosis",
"model",
"organisms",
"cell",
"division",
"proteins",
"gene",
"expression",
"biology",
"mouse",
"molecular",
"biology",
"biochemistry",
"cell",
"biology",
"genetic",
"screens",
"genetics",
"genomics",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] |
2012
|
A Synthetic Interaction Screen Identifies Factors Selectively Required for Proliferation and TERT Transcription in p53-Deficient Human Cancer Cells
|
Human T-cell leukemia virus type 1 ( HTLV-1 ) causes adult T-cell leukemia-lymphoma ( ATL ) and inflammatory diseases . To enhance cell-to-cell transmission of HTLV-1 , the virus increases the number of infected cells in vivo . HTLV-1 bZIP factor ( HBZ ) is constitutively expressed in HTLV-1 infected cells and ATL cells and promotes T-cell proliferation . However , the detailed mechanism by which it does so remains unknown . Here , we show that HBZ enhances the proliferation of expressing T cells after stimulation via the T-cell receptor . HBZ promotes this proliferation by influencing the expression and function of multiple co-inhibitory receptors . HBZ suppresses the expression of BTLA and LAIR-1 in HBZ expressing T cells and ATL cells . Expression of T cell immunoglobulin and ITIM domain ( TIGIT ) and Programmed cell death 1 ( PD-1 ) was enhanced , but their suppressive effect on T-cell proliferation was functionally impaired . HBZ inhibits the co-localization of SHP-2 and PD-1 in T cells , thereby leading to impaired inhibition of T-cell proliferation and suppressed dephosphorylation of ZAP-70 and CD3ζ . HBZ does this by interacting with THEMIS , which associates with Grb2 and SHP-2 . Thus , HBZ interacts with the SHP containing complex , impedes the suppressive signal from PD-1 and TIGIT , and enhances the proliferation of T cells . Although HBZ was present in both the nucleus and the cytoplasm of T cells , HBZ was localized largely in the nucleus by suppressed expression of THEMIS by shRNA . This indicates that THEMIS is responsible for cytoplasmic localization of HBZ in T cells . Since THEMIS is expressed only in T-lineage cells , HBZ mediated inhibition of the suppressive effects of co-inhibitory receptors accounts for how HTLV-1 induces proliferation only of T cells in vivo . This study reveals that HBZ targets co-inhibitory receptors to cause the proliferation of infected cells .
Human T-cell leukemia virus type 1 ( HTLV-1 ) belongs to the delta type retrovirus group , which also includes bovine leukemia virus and HTLV-2 . HTLV-1 causes adult T-cell leukemia-lymphoma ( ATL ) and inflammatory diseases [1–4] . This virus induces clonal proliferation of infected cells to enhance its transmission , since HTLV-1 is transmitted primarily by cell-to-cell contact [5–7] . It has been reported that an increased number of infected cells is correlated with a higher rate of transmission by breast-feeding [8] . Thus , increased numbers of HTLV-1 infected cells are beneficial for the transmission of this virus . HTLV-1 encodes two regulatory genes , tax and rex , and three accessory genes ( p12 , p13 , and p30 ) in the plus strand of the provirus [2] . Another regulatory gene , the HTLV-1 bZIP factor ( HBZ ) gene , is transcribed as an anti-sense transcript [9 , 10] . It has been reported that HTLV-1 infected cells show higher susceptibility to antigenic stimulation . One mechanism of this hypersensitivity is due to Tax . Tax expression under control of the long terminal repeat ( LTR ) results in enhanced responsiveness to stimulation through the T-cell receptor ( TCR ) /CD3 complex [11 , 12] . However , Tax expression is often lost in ATL cells and HTLV-1 infected cells [13–18] . Therefore , it is likely that another mechanism also promotes proliferation of HTLV-1 infected cells–perhaps a mechanism involving HBZ . Indeed , HBZ has been reported to promote proliferation of T cells in vivo and in vitro [19 , 20] . The TCR recognizes cognate antigenic peptides presented by major histocompatibility complex molecules on antigen-presenting cells , and transduces a signal that is modulated by co-stimulatory and co-inhibitory receptors on the T cell [21 , 22] . It has been reported that ATL cells and HTLV-1 infected cells express the co-inhibitory receptors PD-1 and T cell immunoglobulin and ITIM domain ( TIGIT ) on their surfaces [23–25] . Binding of one of these receptors to its ligand sends a suppressive signal through the ITIM or ITSM motif in the cytoplasmic region of the receptor [21] . However , ATL cells and HTLV-1 infected cells proliferate regardless of the higher expression of PD-1 and TIGIT on their surfaces . Until now , it has not been known how the suppressive signal from these co-inhibitory receptors is impaired . In this study , we found that HBZ promotes T-cell proliferation mediated by TCR signaling . As a mechanism , HBZ interferes with the suppressive function of some co-inhibitory receptors and inhibits the expression of others . Thus , impairment of co-inhibitory receptors is a newly discovered mechanism by which HTLV-1 promotes the proliferation of infected T cells .
We have reported that HBZ promotes proliferation of a human T-cell line and HBZ knockdown inhibits proliferation of ATL cell lines [19] . Several mechanisms were identified for proliferation induced by HBZ [20 , 26–31] . However , it remains unknown how HTLV-1 induces T-cell specific proliferation . We generated HBZ transgenic ( HBZ-Tg ) mice , in which HBZ is expressed under the control of the CD4 promoter/enhancer/silencer , so that only CD4+ T cells express HBZ [19 , 32] . We also generated tax transgenic ( tax-Tg ) mice using the same promoter [33] . We isolated CD4+ T cells from HBZ-Tg and tax-Tg mice and evaluated their proliferation upon anti-CD3 stimulation . CD4+ T cells of HBZ-Tg mice proliferated much more than those of non-transgenic ( non-Tg ) mice , and the proliferation of CD4+ T cells was slightly enhanced in tax-Tg mice ( Fig 1A and 1B ) . Co-culture of CD4+ T cells with dendritic cells ( DC ) further enhanced this proliferation ( Fig 1A and 1B ) . However , the difference in proliferation between cells from HBZ-Tg and non-Tg mice was not observed in the presence of anti-CD28 antibody ( 0 . 3 μg/mL ) ( Fig 1C ) , indicating that CD4+ T cells of HBZ-Tg mice are hypersensitive to signaling via the TCR/CD3 complex . To investigate whether the proliferation of CD4+ T cells of HBZ-Tg mice is increased in vivo , we induced experimental allergic encephalomyelitis ( EAE ) in HBZ-Tg and non-Tg mice by immunization with myelin oligodendrocyte glycoprotein ( MOG ) /complete Freund's adjuvant . Although disease severity was not different between HBZ-Tg mice and non-Tg mice ( S1 Fig ) , the number of CD4+ T cells was increased only in the immunized HBZ-Tg mice compared with non-immunized HBZ-Tg and non-Tg mice ( Fig 1D ) , suggesting that HBZ-expressing T cells have higher susceptibility to immune stimulation in vivo . HBZ-Tg did not show higher susceptibility to EAE regardless of impaired Treg function by HBZ [32] . It is speculated that partial inhibition of Treg functions by HBZ is not enough to increase incidence of EAE . It has been reported that HTLV-1 infected cells and ATL cells express both co-stimulatory ( OX40 ) and co-inhibitory receptors ( PD-1 and TIGIT ) on their surfaces [23 , 25 , 34 , 35] . These findings suggest that HTLV-1 influences expression of co-inhibitory and co-stimulatory receptors . Therefore , we analyzed their expressions in HBZ-expressing T cells by real-time RT-PCR . As shown in Fig 2A , the expression of the co-inhibitory receptors TIGIT and PD-1 was enhanced , whereas transcription of other co-inhibitory receptors , BTLA and Lair-1 , was suppressed in HBZ transduced T cells . HBZ suppressed somewhat the transcription of the co-stimulatory receptors CD28 and ICOS but did not influence that of OX40 . In accordance with this finding , flow cytometric analyses showed that in CD4+ T cells from HBZ-Tg mice , cell-surface PD-1 and TIGIT were enhanced , while BTLA and LAIR-1 were decreased ( Fig 2B and 2C ) . HBZ changed the cell-surface expression of co-stimulatory receptors only slightly or not at all ( S2 Fig ) . On the other hand , expression of co-inhibitory receptors did not differ on CD4+ T cells between tax-Tg and non-Tg mice ( S3 Fig ) . Expression of HBZ and tax in these transgenic mice was confirmed by RT-PCR ( S4 Fig ) . To study whether similar changes in levels of these co-stimulatory and co-inhibitory receptors are observed in ATL cells , we analyzed transcription and cell surface expression of these co-receptors . As shown in Fig 3A and 3C , TIGIT transcription and expression were significantly increased in ATL cases . PD-1 expression was upregulated in some ATL cases as reported previously [24] . BTLA transcription and cell-surface expression were not different in ATL cases compared with resting T cells , but suppressed compared with activated T cells ( Fig 3A and 3C ) . Cell-surface expression of LAIR-1 was also suppressed in ATL cells . Transcripts of the HBZ and tax genes were measured by real-time RT-PCR in these cases ( S5 Fig ) . It makes sense that HBZ might decrease the expression of BTLA and LAIR-1 , thus impairing their suppressive function and enhancing the proliferation of infected cells . However , enhanced expression of PD-1 and TIGIT would augment the suppressive function of these co-inhibitory receptors , leading to decreased proliferation of cells . This idea is not consistent with the observation that HBZ enhances proliferation of expressing T cells . Therefore , we speculated that even though HBZ increases the expression of TIGIT and PD-1 , it may inhibit their suppressive function . On the other hand , expression of co-stimulatory receptors , ICOS and OX40 was decreased in ATL cases compared with control activated CD4+ T cells ( S6 Fig ) . The co-inhibitory receptors PD-1 and TIGIT possess ITIM or ITSM domains , and inhibit cell proliferation [21] . As described above , we hypothesized that HBZ may interfere with the T-cell inhibitory function induced by PD-1/PD-L1 and/or TIGIT/CD155 interaction . To study this possibility , we next analyzed the suppressive activity of TIGIT/CD155 interaction in the presence or absence of HBZ . CD4+ T cells were transduced with retroviruses expressing HBZ and stimulated with anti-CD3/CD155 . Fc-coated beads or anti-CD3/control . Fc-coated beads . We then measured proliferation of the cells . As shown in Fig 4A , interaction with anti-CD3/CD155 . Fc-coated beads suppressed the proliferation of CD4+ T cells transduced with empty vector , but not those transduced with HBZ , indicating that HBZ impairs TIGIT/CD155 mediated growth inhibition . Likewise , HBZ interfered with the suppressive effect of PD-1/PD-L1 interaction ( Fig 4A ) . These data suggest that HBZ targets a common molecule ( s ) that is involved in mediating suppressive signals from both PD-1 and TIGIT . Furthermore , suppressive signal through BTLA was also inhibited by the presence of HBZ ( S7 Fig ) . Thus , HBZ not only suppresses BTLA expression but also functionally inhibits suppressive signaling from BTLA . Inhibitory signals through the ITIM and ITSM motifs of PD-1 and TIGIT are mediated by SHP-1 and SHP-2 [21 , 36 , 37] , negative regulators of TCR signaling that dephosphorylate ZAP-70 and CD3ζ and suppress T-cell activation [38 , 39] . We analyzed whether HBZ influences the interaction between the intracytoplasmic region of PD-1 and SHP-2 . To study the interaction of SHP-2 with the cytoplasmic region of PD-1 , we generated a chimeric molecule in which the intracytoplasmic region of human PD-1 was fused to the transmembrane and extracytoplasmic regions of mouse CD28 ( mCD28/hPD1 ) [40] . Pervanadate induces tyrosine phosphorylation of intracellular proteins including PD-1 , thus recruiting SHP-2 to the ITSM motif [40] . After treatment with pervanadate , SHP-2 was recruited to the chimeric molecule ( Fig 4B , lane 2 ) , while HBZ inhibited this interaction ( lane 4 ) . Thus , HBZ hinders recruitment of SHP-2 to the ITSM motif of PD-1 . Binding of PD-L1 induces a transient PD-1-TCR co-localization within microclusters–a co-localization that transiently associates with SHP-2 [41] . After stimulation by pervanadate , PD-1 formed TCR microclusters in Jurkat cells as reported previously ( S8 Fig ) . Next , we analyzed the co-localization of PD-1 and SHP-2 after treatment with pervanadate to observe whether HBZ inhibits the interaction between PD-1 and SHP-2 . As shown in Fig 5A , SHP-2 co-localized with PD-1 in Jurkat cells after pervanadate stimulation ( stimulated Jurkat-mock ) . However , co-localization of these molecules was suppressed by the presence of HBZ ( stimulated Jurkat-HBZ ) . To quantitatively analyze the co-localization of PD-1 and SHP-2 , we visualized PD-1 and SHP-2 using confocal microscopy . Captured raw images were analyzed using ImageJ software with the JACoP plug-in , and Pearson’s correlation coefficient [an index for the relationship of green ( PD-1 ) and red ( SHP-2 ) pixels] was calculated ( Fig 5B ) . Co-localization of PD-1 and SHP-2 in Jurkat mock-transfected cells was minimal without pervanadate stimulation ( the mean correlation coefficient value was 0 . 45 ) , but these two molecules were highly co-localized upon pervanadate stimulation ( correlation coefficient of 0 . 72 ) . This value decreased dramatically , to 0 . 48 , in the presence of HBZ ( Jurkat-HBZ ) , indicating that HBZ strongly interferes with the co-localization of PD-1 and SHP-2 . These data show that HBZ inhibits recruitment of SHP-2 to the cytoplasmic region of PD-1 . As shown above , HBZ inhibits the interaction between SHP-2 and PD-1 . Indeed , phosphorylation of SHP-2 ( Tyr580 ) was decreased in CD4+ T cells of HBZ-Tg mice ( Fig 6A ) and in HBZ-transduced murine primary T cells ( Fig 6B ) . SHP-2 functions to dephosphorylate ZAP-70 and CD3ζ . After induction of phosphorylation by H2O2 [42] , tyrosine phosphorylation of ZAP-70 lasted for a longer time in the presence of HBZ ( Fig 6C ) . Similarly , tyrosine phosphorylation of CD3ζ persisted longer after it was induced by pervanadate ( Fig 6D ) . These data indicate that HBZ interferes with the function of SHP-2 , leading to suppressed dephosphorylation of ZAP-70 and CD3ζ . PD-1 suppresses T-cell proliferation not only by interacting with SHP-1 and SHP-2 , but also by interacting with PKCθ . Threonine phosphorylation ( T538 ) of PKCθ is associated with its activation and IL-2 production by T cells . Signaling via PD-1 inhibits PKCθ T538 phosphorylation [43] . As shown in Fig 6E , HBZ did not enhance phosphorylation of PKCθ T538 , in contrast to ZAP-70 and CD3ζ . These results suggest that HBZ mediated activation of TCR signaling is mainly through inhibition of the tyrosine phosphatase , SHP-2 . This study shows that HBZ inhibits the recruitment of SHP-2 to the cytoplasmic region of PD-1 . However , it remains unknown how HBZ interacts with the complex containing SHP-2 . Recently , THEMIS has been reported to interact with Grb2 and SHP-1 or 2 , and inhibit T-cell activation [44] . We analyzed the interaction of HBZ with these host factors and found that HBZ binds to THEMIS , but not to Grb2 and SHP-2 ( Fig 7A and S9 Fig ) , suggesting that THEMIS is a target of HBZ . Next , we analyzed whether HBZ affects the interaction between THEMIS and Grb2 . We confirmed that THEMIS interacts with Grb2 , and found that this interaction is hindered by the presence of HBZ ( Fig 7B and 7C ) . These data demonstrate that HBZ interacts with THEMIS and partially impairs the association of THEMIS with Grb2 . This interaction of HBZ with the complex containing SHP may hinder recruitment of SHP to the ITSM and ITIM motifs of co-inhibitory receptors such as PD-1 . Next , we analyzed whether HBZ interferes co-localization of PD-1 and THEMIS in the T cells . Stimulation by pervanadate induced co-localization of PD-1 and THEMIS , which was inhibited by HBZ ( Fig 8A ) . Thus , HBZ interacts with THEMIS , which perturbs the complex containing SHP and impairs suppressive signal from PD-1 or TIGIT . To check whether suppressed THEMIS enhances T-cell proliferation through disrupted negative signal , we inhibited THEMIS expression by shRNA and found that suppressed THEMIS expression decreased T-cell proliferation ( S10 Fig ) . It has been reported that proliferation of T cells from THEMIS knockout mice was suppressed [45] , suggesting that THEMIS is also critical for T-cell proliferation in addition to suppressive signaling from co-inhibitory receptors . THEMIS interacts with ITIM or ITSM domain of PD-1 and TIGIT in the cytoplasm whereas it has been reported that HBZ is primarily localized in the nucleus [46] . Therefore , localization of HBZ was analyzed in the presence of THEMIS . As reported previously , THEMIS existed in the cytoplasm ( 50 of 50 cells: 100% ) whereas HBZ was mainly localized in the nucleus of 293T cells ( 67 of 74 cells: 90 . 5% ) ( Fig 8B ) . When both proteins were expressed , HBZ was co-localized with THEMIS in the cytoplasm ( 28 of 79 cells: 35 . 4% ) ( Fig 8B ) . Thus , THEMIS shifted localization of HBZ from nucleus to cytoplasm in 293T cells . When we analyzed localization of HBZ in T cells , HBZ was detected in both the nucleus and the cytoplasm ( Fig 8C ) . Then , we analyzed THEMIS expression in 293T and Jurkat cells , and found that only Jurkat cells expressed THEMIS ( Fig 8D ) , suggesting that THEMIS is responsible for cytoplasmic localization of HBZ . To clarify the localization of HBZ in T cells in detail , we detected HBZ using antibody to the nuclear pore complex , and confirmed that HBZ was present in both nucleus and cytoplasm ( Fig 8E ) . To study whether cytoplasmic localization of HBZ is attributed to THEMIS , we suppressed THEMIS expression using shRNA , and found that HBZ was present largely in the nucleus , suggesting that endogenous THEMIS contributes to changed localization of HBZ from the nucleus to the cytoplasm ( Fig 9 ) .
Co-stimulatory and co-inhibitory receptors control T-cell function and determine T-cell fate after a T cell is stimulated by TCR signaling [21] . In this study , we showed that HBZ enhances the susceptibility of expressing T cells to TCR-mediated signaling by perturbing signaling from co-inhibitory receptors . This study is the first to demonstrate that HBZ targets various co-inhibitory receptors by different mechanisms , and enhances proliferation . Expression of BTLA and LAIR-1 is decreased by HBZ , while HBZ impairs the suppressive function of PD-1 and TIGIT through inhibited recruitment of the SHP-2 containing complex to the cytoplasmic domain of PD-1 . In contrast to BTLA and LAIR-1 , expression of PD-1 and TIGIT are in fact upregulated by HBZ . Why does HBZ enhance expression of TIGIT and PD-1 among co-inhibitory receptors ? Increased TIGIT expression competes with CD226 , a co-stimulatory receptor , for binding with CD155 , resulting in inhibition of T-cell activation [47] . In addition , HBZ suppressed CD226 expression [25] . Furthermore , our previous study indicated that TIGIT expressed on T cells is implicated in immune suppression through enhanced production of IL-10 from T cells and DC by reverse signaling [25] . Since reverse signal from PD-L1 and L2 on DC is also associated with suppressive phenotype of DC and moderate increase in IL-10 expression [48] , PD-1 on T cells is also implicated in immune suppression . DC expresses the TIGIT ligand , CD155 , and PD-1 ligands , PD-L1 and PD-L2 , on the surface . DC-T-cell interaction plays a key role in immune responses to viral infections [49] . We have reported that increased expression of TIGIT on HBZ expressing T cells induces IL-10 production [25] . Furthermore , IL-12p40 production of DC cells was severely impaired in HBZ-Tg mice , which is likely caused by TIGIT on T cells [25 , 47] . Thus HBZ suppresses host immune responses through enhanced PD-1 and TIGIT expression while simultaneously impairing SHP-2 mediated inhibitory signaling from these co-inhibitory receptors . In other words , HBZ modifies the functions of the co-inhibitory receptors PD-1 and TIGIT to allow the virus to evade the host immune system . SHP-1 and 2 form complexes with Grb2 and THEMIS in T cells , and inhibit TCR mediated signaling [44] . Knockdown of THEMIS increased TCR-induced CD3ζ phosphorylation , a phenomenon that resembles the changes caused by HBZ . In this study , we show that HBZ interacts with THEMIS and weakens the interaction between THEMIS and Grb2 . These data suggest that HBZ binding to THEMIS hinders recruitment of this complex to the ITSM motif of PD-1 and thus impedes suppressive signals . THEMIS is expressed only in the T-cell lineage [44 , 50] . Therefore , it is thought that HBZ may not inhibit the co-inhibitory signal in non-T cells . Since the receptor for HTLV-1 is glucose transporter 1 , HTLV-1 infects a variety of cells in vivo [51] . However , only infected T cells proliferate in vivo . Our observation that HBZ binds to THEMIS and impairs the growth-suppressive signal might account for this T-cell specificity of HTLV-1 . TCF-1 and LEF-1 , transcription factors of classical Wnt signaling pathway , are critical for T-cell development in the thymus [52] , and their expressions are suppressed in peripheral memory T cells . We have reported that TCF-1 and LEF-1 inhibit function of Tax , which may critically influence the peripheral T-cell tropism of this virus [53] . It remains unknown how this virus specifically promotes proliferation of infected mature T cells . This study reveals that interaction between HBZ and THEMIS impedes suppressive signal by interfered recruitment of SHP-2 to ITSM motif in T cells . This is thought to be a mechanism of T-cell specificity in HTLV-1 induced proliferation . Thus , HBZ and Tax determine specificity of HTLV-1 infected cells . Analysis of the transcriptomes of 57 ATL cases using RNA-seq identified fusion gene products that contained five CD28-related in-frame fusions ( CTLA4-CD28 or ICOS-CD28 ) that likely induce continuous or prolonged CD28 co-stimulatory signaling [54] . In addition , amplification of CD28 was frequently detected in ATL cases . Thus the CD28 co-stimulatory molecule is a frequent target of somatic changes in ATL cells . As shown in this study , HBZ perturbs co-inhibitory signaling , which enhances proliferation of not only ATL cells but also other HTLV-1 infected cells . Thus , in ATL cells , TCR signaling is a target of both somatic mutation and HBZ . We assume that HBZ perturbs TCR signaling and promotes proliferation beginning in individuals at the carrier state , and then somatic mutations potentiate and exacerbate the proliferative responses . Recently it has been reported that structural variations in the 3’ untranslated region of PD-L1 enhanced PD-L1 expression in various cancer cells [55] . In particular , these structural variations were frequently observed in ATL ( 27% of 49 ATL cases ) . Overexpression of PD-L1 on ATL cells inhibits immune attack from CD8+ T cells through interaction with PD-1 . Since ATL cells also express PD-1 , PD-L1 might suppress the proliferation of ATL cells . However , since HBZ interrupts the suppressive signal from PD-1 through inhibited recruitment of the SHP containing complex to the ITSM motif , ATL cells avoid growth suppression while maintaining the immune suppressive effects of PD-1/PD-L1 interaction on CD8+ T cells . HBZ is primarily localized in the nucleus [46] , and interacts with transcription factors , which include p65 , Smad3 , and c-Jun , and other host factors such as p300 in the nucleus [56–59] . However , it has been reported that HBZ interacts with GADD34 in the cytoplasm , which activates the mammalian target of rapamycin ( mTOR ) signaling [60] . This finding suggests that HBZ also functions in the cytoplasm by interacting with host factors . This study showed that interaction of HBZ with THEMIS changes its localization and HBZ functions in the cytoplasm , and HBZ interferes suppressive function of co-inhibitory receptors . Thus , HBZ exerts the functions in both nucleus and cytoplasm . In this study , we demonstrate that HBZ inhibits suppressive signaling from co-inhibitory receptors by decreased transcription or by inhibition of recruitment of SHP-2 . Furthermore , HBZ enhances the expression of TIGIT and PD-1 , which are associated with immune suppression . Thus , HBZ enables expressing T cells to survive and proliferate in vivo by utilizing and modifying the functions of co-inhibitory receptors .
C57BL/6J mice were purchased from CLEA Japan . Transgenic mice expressing HBZ ( HBZ-Tg mice ) under the control of the murine CD4-specific promoter/enhancer/silencer have been described previously [19] . All HBZ-Tg mice were heterozygotes for the transgene . Transgenic mice expressing tax ( tax-Tg mice ) under the control of the same promoter were generated as reported [33] . Jurkat cell line was provided by Dr . S . Sakaguchi ( Osaka University , Japan ) . Jurkat cell lines stably expressing the spliced form of HBZ ( Jurkat-HBZ ) and control ( Jurkat-mock ) cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum ( FBS ) and 1 mg/mL G418 ( Nacalai Tesque , Kyoto , Japan ) [61] . The 293T cell line was purchased from ATCC ( Manassas , VA , USA ) and cultured in Dulbecco’s modified Eagle medium ( DMEM ) supplemented with 10% FBS . The 293FT cell line was purchased from Life Technologies and cultured in DMEM supplemented with 10% FBS and 0 . 5 mg/mL G418 . The packaging cell line , Plat-E , was provided by Dr . T . Kitamura ( Institute of Medical Science , The University of Tokyo , Japan ) and cultured in DMEM containing 10% FBS , 10 μg/mL blasticidin and 1 μg/mL puromycin . These cell lines were grown at 37°C under a 5% CO2 atmosphere . Peripheral blood mononuclear cells ( PBMCs ) of ATL patients and healthy donors were collected by Ficoll-Paque PLUS ( GE Healthcare , Little Chalfont , UK ) . CD4+ T cells of healthy donors were isolated by Human CD4+ T Cell Enrichment Cocktail ( STEMCELL Technologies , Vancouver , Canada ) according to the manufacturer’s instructions . To obtain activated CD4+ T cells , CD4+ T cells were stimulated by 10 μg/mL phytohemagglutinin ( PHA ) ( Sigma-Aldrich , St . Louis , MO , USA ) for three days . Animal experiments were performed in strict accordance with the Japanese animal welfare bodies ( Law No . 105 dated 19 October 1973 , modified on 2 June 2006 ) , and the Regulation on Animal Experimentation at Kyoto University . The protocol was approved by the Institutional Animal Research Committee of Kyoto University ( permit numbers D13-02 , D14-02 , D15-02 , and A10-3 ) . Experiments using clinical samples were conducted according to the principles expressed in the Declaration of Helsinki , and approved by the Institutional Review Board of Kyoto University ( permit number G310 ) . ATL patients and healthy blood donors provided written informed consent for the collection of samples and subsequent analysis . pMX-IG , pMX-HBZ-IG and pMX-BTLA-IG were used for retrovirus production . The coding sequence of BTLA was amplified from cDNA of a wild type C57BL/6J mouse , and subcloned into pMX-IG . The HBZ expression vector , pcDNA3 . 1 HBZ-mycHis , was described previously [58] . The SHP-2 expression vector , pSP65SRa-SHP2-Flag , was kindly given by Dr . M . Hatakeyama ( The University of Tokyo , Japan ) . The entire coding regions of Grb2 and THEMIS were amplified from cDNA prepared from resting human PBMCs or Jurkat cells . These PCR fragments were subcloned into pCMV-HA ( Clontech Laboratories , Palo Alto , CA , USA ) or pCAGGS-PA . The resulting plasmids were designated pCMV-HA-Grb2 and pCAGGS-PA-THEMIS , and they express HA ( YPYDVPDYA ) -tagged Grb2 and PA ( GVAMPGAEDDVV ) -tagged THEMIS , respectively . For generation of the chimeric mCD28/human PD-1 ( hPD-1 ) expression vector , the mCD28 extracellular and transmembrane domains ( bases 9–617 of NM_007642 ) were amplified from cDNA prepared from stimulated murine T cells , and the hPD-1 intracellular domain ( bases 645–935 of NM_005018 ) was amplified from cDNA prepared from PHA-stimulated human PBMCs . These fragments were ligated and the resultant fragment was substituted for the GFP cording region of pMX-HBZ-IRES-GFP . Anti-CD3 antibody ( 145-2X11 , R&D systems , Minneapolis , MN , USA ) together with recombinant mouse CD155 . Fc or the control . Fc ( Sino Biological , Beijing , China ) or PD-L1 . Fc or the control . Fc ( R&D systems ) or HVEM . Fc ( R&D systems ) were covalently attached to Dynabeads M450 Tosylactivated ( Invitrogen , Thermo Fisher Scientific , Waltham , MA , USA ) . Anti-CD3 antibody together with control . Fc was used for the control . For each 107 beads , 1 μg of anti-CD3 antibody ( 20% of total protein ) and 4 μg of CD155 . Fc , PD-L1 . Fc , HVEM . Fc or control . Fc ( 80% ) were used . The PiggyBac-based shRNA expression vector , pB-CMV-GreenPuro-H1 ( System Biosciences , Palo Alto , CA , USA ) , containing shRNA against THEMIS or luciferase ( control ) , was introduced into Jurkat cells together with PiggyBac transposase expression vector by using Neon Transfection System ( Invitrogen ) . Target sequences of each shRNA are shown in S1 Table . To measure the proliferation of CD4+ T cells of HBZ-Tg mice , we isolated murine splenic CD4+ T cells using the CD4 T Lymphocyte Enrichment Set ( BD Biosciences , San Jose , CA , USA ) . Murine splenic dendritic cells were isolated from collagenase-digested low-density cells using the Dendritic Cell Enrichment Set ( BD Biosciences ) . Purified CD4+ T cells were labeled with 5- ( and-6 ) -carboxyfluorescein diacetate succinimidyl ester ( CFSE , Molecular Probes , Thermo Fisher Scientific , Waltham , MA , USA ) . Labeled CD4+ T cells of HBZ-Tg , non-Tg and tax-Tg mice ( 2×105 cells/well ) were cultured with or without dendritic cells ( 1×104 cells/well ) from non-Tg mice for three days with soluble anti-CD3 antibody ( 30 ng/mL ) stimulation in round-bottomed 96-well plates . CFSE dilution was analyzed by flow cytometry . For TIGIT/CD155 and PD-1/PD-L1 proliferation assays , HBZ or empty vector transduced cells ( see below ) were labeled with CellTrace Violet ( Invitrogen ) and stimulated with anti-CD3/CD155 . Fc or anti-CD3/PD-L1 . Fc or anti-CD3/control . Fc-coated beads at a bead-to-cell ratio of 1:1 for three days . Dye dilution was analyzed by flow cytometry . Transfection of the packaging cell line , Plat-E , was performed as reported [62] . Murine CD4+ T cells were activated by immobilized anti-CD3 ( 1 μg/mL ) and soluble anti-CD28 ( 0 . 1 μg/mL ) in 12-well plates . After 24 hours , activated T cells were transduced with virus supernatant and 4 μg/mL polybrene , and centrifuged at 3 , 000 rpm for 60 min . Cells were subsequently cultured for 48 hours . Total RNA was isolated using Trizol Reagent ( Invitrogen ) and treated with DNase I to remove the genomic DNA . cDNAs were synthesized from 1 μg of total RNA using random primer and SuperScript III or IV reverse transcriptase according to the manufacturer’s instructions ( Invitrogen ) . mRNA expression was analyzed by real-time PCR using FastStart Universal SYBR Green Master ( Roche Diagnostics , Basel , Switzerland ) and the StepOnePlus Real-Time PCR System ( Applied Biosystems , Thermo Fisher Scientific , Waltham , MA , USA ) according to the manufacturer’s instructions . Primers used in this study are shown in S1 Table . The following antibodies were used for flow cytometric analyses . Anti-mCD4 ( GK1 . 5 ) , mTIGIT ( IG9 ) , mCD28 ( E18 ) , mICOS ( C398 . 4A ) , mOX40 ( OX-86 ) , hCD4 ( RPA-T4 ) , hPD-1 ( 29F . 1A12 ) , hBTLA ( MIH26 ) , hLAIR-1 ( NKTA255 ) , hCD28 ( CD28 . 2 ) , hICOS ( C398 . 4A ) and hOX40 ( Ber-ACT35 ) antibodies were all purchased from BioLegend ( San Diego , CA , USA ) . Anti-mBTLA ( 6F7 ) , mLAIR-1 ( 113 ) and hTIGIT ( MBSA43 ) antibodies were purchased from eBioscience ( San Diego , CA , USA ) . Anti-mPD-1 ( J43 ) was from BD Pharmingen ( BD Biosciences ) . Anti-phospho-SHP-2 ( Tyr580 ) antibody was from Cell Signaling Technology ( Danvers , MA , USA ) . Anti-mouse IgG1 , mouse IgG2b and rat IgG1 ( BioLegend ) , Armenian hamster IgG ( eBiosciences ) and mouse IgG2a ( BD Pharmingen ) were purchased for isotype controls . For detection of SHP-2 phosphorylation , cells were permeabilized using BD Phosflow perm buffer II ( BD Biosciences ) according to the manufacturer’s instructions . Flow cytometric analysis was carried out using a FACSVerse with FACSuite software ( BD Biosciences ) and FlowJo ( TreeStar , Ashland , OR , USA ) . For the induction of EAE , five to seven-week old HBZ-Tg or control non-transgenic ( non-Tg ) C57BL/6J mice were immunized with an emulsion containing MOG peptide 35–55 ( MEVGWYRSPFSRVVHLYRNGK ) . The emulsion was prepared by sonication , mixing 1 mL of MOG solution ( 1 mg/mL in PBS ) with 1 mL of complete Freund’s adjuvant ( Difco Laboratories , Detroit , MI , USA ) containing desiccated Mycobacterium butyricum . The emulsion was injected subcutaneously in the area near the axillary lymph nodes and on both sides at the base of the tale of each mouse ( 50 μL/site , a total of 4 sites/mouse ) . On days 0 and 2 post-immunization , mice were injected intraperitoneally with 50 μL of a Pertussis toxin ( Kaketsuken , Kumamoto , Japan ) solution ( 4 μg/mL ) . Thereafter , mice were monitored daily for clinical signs of encephalomyelitis . A clinical score was assigned according to the following criteria: 0 , no symptoms; 1 , mild limp tail; 1 . 5 , limp tail; 2 , unilateral hind limb weakness or abnormal gait; 2 . 5 , unilateral hind limb paralysis or bilateral hind limb weakness; 3 , paraplegia; 3 . 5 , unilateral fore limb weakness , with paraplegia; 4 , unilateral fore limb paralysis or bilateral fore limb weakness; 4 . 5 , bilateral fore limb paralysis; 5 , moribund or dead . For the immunoprecipitation studies of SHP-2 and PD-1 , 293FT cells were transfected with the indicated expression vectors using Lipofectamine LTX ( Invitrogen ) according to manufacturer’s instructions . After 48 hours , cells were stimulated with 0 . 1 mM pervanadate solution for 5 min . The cell lysates were immunoprecipitated for 60 min at 4°C with 5 μg of anti-mCD28 ( 37 . 51 ) , and immune complexes were incubated with Protein G-Sepharose ( GE Healthcare ) for 60 min at 4°C . For the immunoprecipitation studies of THEMIS , Grb2 , SHP-2 and HBZ , 293FT cells were transfected as described above . After 48 hours , cells were stimulated with H2O2 for 5 min . The cell lysates were immunoprecipitated with 20 μg of anti-PA ( NZ-1 ) , anti-HA ( HA-7 ) or anti-Flag ( M2 ) antibodies , and immune complexes were incubated as described above . Normal mouse and rat IgG ( Santa Cruz Biotechnology , Dallas , TX , USA ) were used as controls . The following antibodies were used for immunoblotting: anti-phospho-SHP-2 ( Tyr580 ) , phospho-ZAP-70 ( Tyr319 and Tyr493 ) , ZAP-70 ( 99F2 ) , phospho-PKCθ ( Thr538 ) and PKCθ ( E1I7Y ) antibodies were purchased from Cell Signaling Technology . Anti-phosphor-CD3ζ ( Tyr83 ) , CD3ζ and anti-THEMIS were purchased from Abcam ( Cambridge , UK ) . Anti-PA ( NZ-1 ) was purchased from Wako , Osaka , Japan . Anti-Flag-HRP ( M2 ) , HA-HRP ( HA7 ) , Myc ( 9E10 ) , and tubulin ( DM1A ) antibodies were purchased from Sigma-Aldrich . Anti-mouse IgG-HRP , rabbit IgG-HRP and rat IgG-HRP antibodies were purchased from GE Healthcare . Mouse anti-HBZ monoclonal antibody ( clone 1A10 ) was generated by immunizing C57BL/6 with using keyhole limpet hemocyanin ( KLH ) -conjugated HBZ peptide 97–133 ( CKQIAEYLKRKEEEKARRRRRAEKKAADVARRKQEEQE ) . To detect co-localizations of PD-1 with SHP-2 , THEMIS or TCR , Jurkat-mock or Jurkat-HBZ cells were stimulated with 0 . 2 mM pervanadate solution for 2 min at 37°C . To evaluate the effect of THEMIS on the localization of HBZ , THEMIS-knocked down ( KD ) and control ( luciferase KD ) Jurkat cells were transfected with pcDNA 3 . 1 HBZ-mycHis or empty vector . The cells were washed with PBS and placed on MAS-coated glass slides ( Matsunami Glass , Osaka , Japan ) . To detect HBZ , 293T cells cultured on type I collagen ( Cellmatrix , Nitta Gelatin , Osaka , Japan ) -coated coverslips , were transfected with pcDNA 3 . 1 HBZ-mycHis and/or pCAGGS-PA-THEMIS . The cells were fixed with 4% paraformaldehyde for 15 min , permeabilized with 0 . 2% Triton X-100 for 15 min , and blocked by incubation in 5% donkey serum ( Jackson ImmunoResearch , West Grove , PA , USA ) for 60 min . For immunostaining , the cells were incubated with anti-SHP-2 ( sc-280 ) , anti-PD-1 ( sc-10297 ) , anti-TCR β ( sc-5277 ) ( all Santa Cruz Biotechnology ) , anti-THEMIS ( ab126771 ) , anti-Nuclear Pore Complex Proteins ( ab24609 ) ( all Abcam ) , anti-myc ( 9E10 , Sigma-Aldrich or Abcam ) or anti-PA ( NZ-1 ) antibodies for 60 min , followed by incubation with Alexa Fluor 488-conjugated donkey anti-goat IgG , Alexa Fluor 488-conjugated donkey anti-rat IgG , Alexa Fluor 594-conjugated donkey anti-mouse IgG , Alexa Fluor 594-conjugated donkey anti-rat IgG , Alexa Fluor 647-conjugated donkey anti-mouse IgG or Alexa Fluor 647-conjugated donkey anti-rabbit IgG antibodies ( all Invitrogen ) , or DyLight 405-conjugated donkey anti-mouse IgG ( Jackson ImmunoResearch ) for 30 min . The stained cells were mounted with ProLong Gold Antifade Reagent or ProLong Gold Antifade Reagent with DAPI ( all Molecular Probes , Thermo Fisher Scientific ) , imaged using an FV1000 confocal microscope ( Olympus , Tokyo , Japan ) or a Leica TCS SP8 ( Leica Microsystems , Wetzlar , Germany ) , and analyzed with ImageJ . For Figs 1 , 2 , 7 , S2 and S10 , statistical significance was determined by the two-tailed unpaired Student’s t-test . For Figs 3 , 5 and S6 , statistical analysis was performed using the one-way ANOVA with Tukey’s post hoc test ( GraphPad Prism , GraphPad Software , La Jolla , CA , USA ) . Asterisks indicate the statistical significance as follows: *P < 0 . 05; **P < 0 . 01; ***P < 0 . 001; n . s . , not significant .
|
Since HTLV-1 infects only through cell-to-cell transmission , increasing the number of infected cells is critical for transmission of HTLV-1 . Proliferation of HTLV-1 infected cells is critical for development of leukemia and inflammatory diseases . In this study , we showed that HBZ promotes the proliferation of infected cells by targeting co-inhibitory receptors . Paradoxically , HBZ enhances the expression of the co-inhibitory receptors TIGIT and PD-1 . We found that HBZ concurrently hampers the growth-inhibitory signal of TIGIT and PD-1 , thereby leading to the enhanced proliferation of HTLV-1 infected cells in vivo . HBZ does this by interacting with THEMIS , which is expressed only in T cells . It is known that HTLV-1 infects different types of cells but increases only T cells . Functional impairment of co-inhibitory receptors by interaction of HBZ with THEMIS is a mechanism how HTLV-1 specifically induces proliferation of T cells .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"flow",
"cytometry",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"cd",
"coreceptors",
"pathogens",
"immunology",
"microbiology",
"retroviruses",
"viruses",
"rna",
"viruses",
"immunoprecipitation",
"tcr",
"signaling",
"cascade",
"coreceptors",
"antibodies",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"spectrum",
"analysis",
"techniques",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"htlv-1",
"t",
"cells",
"microbial",
"pathogens",
"precipitation",
"techniques",
"spectrophotometry",
"biochemistry",
"signal",
"transduction",
"t",
"cell",
"receptors",
"cytophotometry",
"cell",
"biology",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"immune",
"receptors",
"cell",
"signaling",
"organisms",
"signaling",
"cascades"
] |
2017
|
HTLV-1 bZIP Factor Enhances T-Cell Proliferation by Impeding the Suppressive Signaling of Co-inhibitory Receptors
|
HSPB7 is a member of the small heat-shock protein ( HSPB ) family and is expressed in the cardiomyocytes from cardiogenesis onwards . A dramatic increase in HSPB7 is detected in the heart and blood plasma immediately after myocardial infarction . Additionally , several single-nucleotide polymorphisms of HSPB7 have been identified to be associated with heart failure caused by cardiomyopathy in human patients . Although a recent study has shown that HSPB7 is required for maintaining myofiber structure in skeletal muscle , its molecular and physiological functions in the heart remain unclear . In the present study , we generated a cardiac-specific inducible HSPB7 knockout mouse and demonstrated that the loss of HSPB7 in cardiomyocytes results in rapid heart failure and sudden death . The electrocardiogram showed cardiac arrhythmia with abnormal conduction in the HSPB7 mutant mice before death . In HSPB7 CKO cardiomyocytes , no significant defect was detected in the organization of contractile proteins in sarcomeres , but a severe structural disruption was observed in the intercalated discs . The expression of connexin 43 , a gap-junction protein located at the intercalated discs , was downregulated in HSPB7 knockout cardiomyocytes . Mislocalization of desmoplakin , and N-cadherin , the intercalated disc proteins , was also observed in the HSPB7 CKO hearts . Furthermore , filamin C , the interaction protein of HSPB7 , was upregulated and aggregated in HSPB7 mutant cardiomyocytes . In conclusion , our findings characterize HSPB7 as an intercalated disc protein and suggest it has an essential role in maintaining intercalated disc integrity and conduction function in the adult heart .
HSPB7 , also known as cardiovascular heat-shock protein ( cvHsp ) , is a member of the small heat-shock protein ( sHSP or HSPB in mammals ) family that shares a conserved α-crystallin domain in the C-terminal region [1 , 2] . The sHSPs characteristically function as ATP-independent molecular chaperones assisting intracellular protein assembly and cytoskeleton formation under normal conditions and suppressing the aggregation of denaturing proteins in resistance to stress [3] . Many sHSPs are expressed in cardiac and skeletal muscles [4 , 5] and the phenotypes of their mutations are seen in muscle diseases . For example , mutation in HSPB5 was found to induce desmin-related skeletal muscle myopathy and cardiomyopathy [6 , 7] . Previous studies have also demonstrated that HSPB2 , HSPB5 , and HSPB6 can protect against myocardial ischemia-reperfusion injury and suppress pressure overload cardiac hypertrophy [8–12] . Consequently , these studies suggest that sHSPs are crucial in protecting striated muscles from damage caused by stress and injury . HSPB7 is the most highly expressed sHSP gene in the heart [1] . Increasing expression of HSPB7 can be detected in the monocrotaline-induced hypertrophic right ventricle [1] , aging skeletal muscle [13] , and dystrophin-deficient MDX diaphragm in mice [14] . Despite being the most potent suppressor of sHSP against the aggregation of the mutated huntingtin protein [15] , HSPB7 was reported to be a potential early biomarker of myocardial infarction and an independent risk factor for acute coronary syndrome [16] . Overexpression of HSPB7 can reduce the amount of tachypacing-induced F-actin stress fibers through attenuation of the RhoA-GTPase pathway [17] . Recent studies have reported that an intronic single-nucleotide polymorphism ( SNP ) rs1739843 in HSPB7 is highly associated with heart failure ( HF ) [18–20] , dilated cardiomyopathy ( DCM ) [21] , and idiopathic DCM [22] in human patients . Although rs1739743 and another 11 additional HSPB7 SNPs were further confirmed to be associated with heart failure , the additional SNPs were also found to be intronic or synonymous . As such , HSPB7 SNPs are predicted to have no effect on its protein sequence or function , suggesting a role as a marker for the position of a genetically linked functional variant located outside HSPB7 [23] . However , all the previous findings still imply the role of HSPB7 in cardiac pathogenesis . Recently , the studies conducted using gene knockdown in zebrafish demonstrated that HSPB7 is required for the formation of the left-right axis and cardiac morphogenesis [24 , 25] . Although these findings demonstrate the essential role of HSPB7 in cardiac development and functional maintenance , the functions of HSPB7 in adult heart still remain unclear . Using an HSPB7 inducible-conditional knockout ( CKO ) model approach , here we demonstrate that the loss of HSPB7 leads to cardiomyopathy and arrhythmic sudden death . Lack of HSPB7 causes the disruption of the intercalated disc ( ID ) structure , resulting in abnormal localization of ID component proteins and defect of the cardiac conduction function . Furthermore , we found that filamin C ( FLNC ) , the interaction protein of HSPB7 , was mislocalized and aggregated in HSPB7 CKO cardiomyocytes . Thus , our results suggest that HSPB7 acts as a novel cardiac ID protein to maintain the structural integrity of the ID and the cardiac functions of the adult heart .
To explore the functional role of HSPB7 in the heart , we first analyzed the expression of HSPB7 in the heart . Immunoblot analysis revealed that HSPB7 is expressed in the heart from embryonic day 14 . 5 ( E14 . 5 ) to postnatal day 28 ( P28 ) ( Fig 1A ) in multiple forms with different molecular masses ( arrows in Fig 1A ) . The subcellular localization of HSPB7 was determined by confocal fluorescence microscopy in longitudinal sections of the adult mouse heart . Double labeling with sarcomere markers , α-actinin as the Z-line , myomesin as the M-line , and cardiac-actin as the I-band showed that HSPB7 is expressed adjacent to the Z-line ( Fig 1B ) . Localization of HSPB7 at the IDs was verified by co-staining with N-cadherin ( adherens junction ) , desmoplakin ( desmosomes ) , and connexin 43 antibody ( gap junction; Fig 1B ) . HSPB7 is present as a diffusion pattern in bands of regular periodicity , and is not completely colocalized with N-cadherin at E18 . 5 ( Fig 1C ) . After birth , bright patches of HSPB7 staining were observed at P3 and appear to have a compact pattern of adherens junction localization at P14 , which is indicated by colocalization with N-cadherin . In the adult heart , HSPB7 staining was seen as faint striations , mostly concentrated at the ends of cardiomyocytes . Our results indicate that HSPB7 is highly colocalized with N-cadherin during the assembly and maturation of IDs , suggesting that HSPB7 may be involved in organizing and maintaining the cardiac cytoarchitecture . To elucidate the function of HSPB7 in the heart , a tamoxifen-induced CKO mouse line ( MCM/HSPB7Flox/Flox ) was established using HSPB7Flox/Flox intercrossed with MCM mice [26] . HSPB7 CKO and littermates ( 8- to 10-week-old ) including HSPB7Flox/Flox and MCM mice were administered with tamoxifen for four consecutive days . Immunoblot analysis of heart lysates revealed a 75% and 95% reduction in HSPB7 protein levels in the hearts of CKO mice compared with their control HSPB7Flox/Flox littermates ( n = 4; Fig 2A and 2B ) at d4 and d7 after tamoxifen administration , respectively . Immunofluorescence staining and confocal microscopy also revealed a dramatic decrease in HSPB7 at the intercalated discs and sarcomeres at d7 after tamoxifen administration in CKO mice ( Fig 2C ) . The CKO mice displayed a moderate increase in heart weight/body weight ratio compared with their HSPB7Flox/Flox littermates ( n = 8; Fig 2D ) . Notably , we found that ablation of HSPB7 in the cardiomyocytes led to rapid mouse death within 12 days ( n = 11 ) . By contrast , tamoxifen-treated HSPB7Flox/Flox ( n = 10 ) or MCM control mice ( n = 11 ) appeared healthy throughout the course of tamoxifen treatment ( Fig 2E ) . Histological analysis showed slight inflammatory infiltrate , cardiomyocyte disarray , and enlarged myocytes with hyperchromatic nuclei in the CKO myocardium . However , fibrosis was not detected in the mutant heart by using Masson’s trichrome stain ( Fig 2F ) . To further explore the functional pathological phenotype of the HSPB7 CKO mice in more detail , we then analyzed cardiac function through echocardiography . Echocardiographic analysis showed that the contractility of the CKO mice was significantly impaired compared with the control mice . We observed a significant reduction in LV fractional shortening ( %FS; 29 . 3% ± 4 . 0 for HSPB7Flox/Flox versus 15 . 6% ± 6 . 7 for CKO ) and ejection fraction ( %EF; 63 . 0% ± 4 . 8 for HSPB7Flox/Flox versus 38 . 1% ± 14 . 5 for CKO ) in CKO mutants compared with controls ( n = 5; Fig 3 ) at d7 after tamoxifen administration . These results imply that ablation of HSPB7 in the myocardium leads to cardiomyopathy , HF , and sudden death . Given the cardiomyopathy and sudden death observed in HSPB7 CKO mice , we speculated that the normal electrophysiological activities were disturbed in the HSPB7 mutant hearts . For routine monitoring of cardiac activity during anesthesia , we used the ECG , which revealed an abnormality in the HSPB7 CKO hearts ( Fig 4A–4D ) . ECG measurements in 8-week-old mice revealed no difference in QRS duration , QT interval , or PQ interval between HSPB7Flox/Flox and MCM/HSPB7Flox/Flox mice ( Fig 4A and 4C ) . By d7 after tamoxifen administration , although the conduction of the electrical activity from the atrium to ventricle of the CKO mice was normal ( PR interval ) ( 53 . 6 ± 6 ms vs . 54 . 3 ± 6 ms; Fig 4B and 4D ) , the depolarization and repolarization of the CKO ventricle were disturbed . This is reflected in the prolonged QRS complex ( 11 . 4 ± 0 . 8 ms vs . 16 . 4 ± 2 . 1 ms ) and QT interval ( 20 . 0 ± 1 . 26 ms vs . 33 . 3 ± 8 . 5 ms ) after induction of the CKO ( Fig 4B and 4D ) . To further investigate the nature of sudden cardiac death and evaluate possible underlying ECG abnormalities , miniaturized telemetric transmitter devices were implanted in 8- to 10-week-old MCM , HSPB7Flox/Flox , and HSPB7Flox/Flox mice to record their cardiac rhythm [27] ( Fig 4E ) . We recorded the time interval for 2 hours before and at d4 , d7 , and d14 after tamoxifen administration to HSPB7 CKO and control mice . The HSPB7 CKO mice were continuously recorded at d7 after tamoxifen administration until death . All the control mice exhibited normal sinus rhythms , with no evidence of ventricular ectopy . We also observed ST segment abnormalities in HSPB7 CKO mice similar to the result of surface ECG recordings at d7 after tamoxifen administration ( S1 Fig ) . In the HSPB7 CKO mice during the continuous recording period , we captured the abrupt onset of spontaneous ventricular tachyarrhythmia , confirming that the death was arrhythmic . To examine the myofibril organization and cell–cell contacts at the ultrastructural level , transmission electron micrograph ( TEM ) analysis was performed on control and CKO hearts ( Fig 5 ) at d7 after tamoxifen administration . Intercalated disc structures were visible in the HSPB7Flox/Flox hearts , with clear adherens junctions and desmosomes represented by submembranous electron dense material adjacent to the intercellular space between the myocytes . By contrast , the structures of the mutant IDs were highly convoluted and disorganized ( Fig 5A ) . Higher-magnification images revealed abnormal adherens junctions ( black arrowhead in Fig 5A ) and desmosomes ( white arrowhead in Fig 5A ) with widened gaps at the IDs of HSPB7 mutant cardiomyocytes . The ultrastructure of the sarcomeres at the center part of the mutant cardiomyocytes was slightly distorted , showing loose actin filaments ( black arrowhead in Fig 5B ) and wider , less dense Z-lines ( white arrowhead in Fig 5B ) compared with the controls . The TEM images of the sarcomeres proximal to the intercalated discs showed an abnormal Z-line ( black arrow in Fig 5A ) , disrupting filaments ( asterisk in Fig 5A ) , and lacunae spaces at the sites of myofibril attachment at the IDs ( white arrows in Fig 5A ) . These sarcomere defects in the HSPB7 CKO myocardium presumably reflect the lack of myofibril anchorage at the plasma membrane . Because HSPB7 is expressed at both the IDs and adjacent to the Z-line in the adult heart , we next examined whether the loss of HSPB7 affects the cell–cell junctions of the IDs and the sarcomeric apparatus . The components of the ID , desmoplakin ( a cytoplasmic desmosomal protein ) , N-cadherin ( an adherens junction protein ) and connexin 43 ( a gap junction protein ) were examined through immunofluorescence staining and confocal microscopy ( Fig 6A ) . In HSPB7 CKO hearts , the cytoplasmic localization of desmoplakin and N-cadherin were observed in the HSPB7 CKO heart ( Fig 6A ) but not in the HSPB7Flox/Flox control heart . Moreover , depletion of HSPB7 in the IDs resulted in a significant decrease in connexin 43 in the myocardium ( Fig 6A ) . Immunoblot blot analysis further confirmed a reduction in connexin 43 in the HSPB7 CKO heart compared with the HSPB7Flox/Flox control heart ( Fig 6B and 6C ) , whereas N-cadherin and desmoplakin levels remained unchanged in the mutant hearts ( Fig 6B and 6C ) . Furthermore , the expression pattern of vinculin ( a costamere marker ) was characterized relatively normally ( Fig 6A ) in HSPB7 CKO cardiomyocytes . Confocal microscopy revealed the normally striated structure of the Z-line ( α-actinin ) and the M-line ( myomesin ) ( S2 Fig ) , consistent with the results of the TEM . Likewise , immunoblot analysis also showed the same expression levels of sarcomeric protein ( Fig 6B and 6C ) in the HSPB7 CKO hearts . Taken together , these data support the notion that HSPB7 is required to maintain the ID structure including the components of gap junctions , desmosomes , and adherens junctions . To further identify the possible functional target ( s ) affected by HSPB7 loss , we examined the interactions of ID component proteins with HSPB7 by co-immunoprecipitation . Interestingly , only FLNC ( but not N-cadherin , desmoplakin , or connexin 43 ) was found to interact with HSPB7 in heart lysate ( S3 Fig ) . FLNC has been found to be the interaction protein of HSPB7 , and the loss of HSPB7 in skeletal muscles can cause progressive myopathy with FLNC aggregation [28] . The confocal microscopy showed that HSPB7 mainly colocalized with FLNC adjacent to the Z-line and at IDs in the cardiomyocytes ( S4 Fig ) . Furthermore , a significant increase in FLNC expression was detected in the HSPB7 CKO heart at d7 after tamoxifen administration ( Fig 7A and 7B ) and is consistent with the immunoblotting analysis in the supernatant and pellet fractions of heart lysates showing the upregulation of FLNC protein expression ( Fig 7C and 7D ) . Furthermore , along with the FLNC aggregation detected in the myocardium ( arrowheads in Fig 7A ) , an accumulation of extracellular matrix ( WGA staining in Fig 7A and 7B ) was also observed in the myocardium of the HSPB7 CKO mice . To understand whether the aggregation of FLNC is the cause of the disruption of ID structure , immunoblotting analysis was performed in HSPB7 CKO mice at d4 after tamoxifen administration ( S5 Fig ) . The results showed that the downregulation of connexin 43 occurred before the upregulation of FLNC in the HSPB7 CKO heart . These results suggest that the reduction of connexin 43 may not be caused by the overexpression or aggregation of FLNC . Additionally , double staining of FLNC with desmoplakin or N-cadherin showed that the mislocalization of desmoplakin or N-cadherin ( arrowheads in S6 Fig ) and FLNC aggregation ( arrows in S6 Fig ) did not always occur at the same cardiomyocytes of the HSPB7 CKO heart at d7 . Our findings indicate that the overexpression or aggregation of FLNC protein may not be the direct cause for the disruption of ID structure in HSPB7 CKO cardiomyocytes . We next examined whether cell integrity was affected by the loss of HSPB7 in the heart [29 , 30] , since the sarcolemma disruption was observed in the HSPB7 skeletal muscle specific CKO mouse . Using EBD uptake analysis , we found that the HSPB7 CKO heart presented blue coloration compared with the control . Consistent with our previous study [28] , a high uptake of EBD ( red ) was detected in the HSPB7 mutant cardiomyocytes ( S7 Fig ) . Tamoxifen in αMHC-MerCreMer mice could induce a DNA damage response leading to HF and death [31] . To exclude the possibility that the phenotype of HSPB7 CKO resulted from tamoxifen toxicity , we performed gene elimination through direct intramyocardial injection with the Adeno-Cre virus in HSPB7Flox/Flox mice . The expression patterns of ID components and FLNC were evaluated at d8 after Adeno-Cre virus injection by immunofluorescence staining . Adeno-Cre virus treatment efficiently reduced HSPB7 expression at the injected region of the heart ( the upper Cre region shown in S8 Fig ) . As shown in the immunofluorescence analysis , HSPB7 was eliminated , and a decrease in expression of connexin 43 was observed in the HSPB7-depleted regions of the Adeno-Cre virus-injected heart . Additionally , high magnification confocal images of these regions showed that the mislocalization of desmoplakin ( Fig 8A ) and N-cadherin ( Fig 8B ) occurred in HSPB7-depleted cardiomyocytes . Consistent with the results for the HSPB7 CKO heart , confocal fluorescence microscopy revealed that the immunoreactivity of FLNC significantly increased in HSPB7 depleted regions of the Adeno-Cre virus-injected heart . Overall , our results indicate a direct functional and cell-autonomous role for HSPB7 in maintaining FLNC stability and ID structure in cardiomyocytes .
In this study , we have shown that HSPB7 plays an essential role in maintaining ID integrity to prevent cardiac arrhythmogenic failure . We demonstrated the dynamic expression and subcellular location of HSPB7 in cardiac muscle from the embryonic stage to adulthood . We found that HSPB7 is highly colocalized with N-cadherin during the assembly and maturation of IDs . Importantly , we demonstrated that the ablation of HSPB7 in adult mouse hearts leads to ( i ) the disruption of ID structure with distorted expression and location of ID components , ( ii ) defects in myofibrillar organization and membrane integrity in cardiomyocytes , and ( iii ) the development of abnormal conductive activity with arrhythmic sudden death . The IDs are an indispensable structure that connect neighboring cardiomyocytes , which is essential for electric , mechanical , and signaling communication between adjacent cells [32] . Loss of ID components , such as plakoglobin [30] and N-cadherin [33] causes cardiac arrhythmic death with distortion of ID structures and downregulation of ID protein in mice . A previous study demonstrated that dysregulation and mislocalization of cadherin may dissipate the contractile force across the plasma membrane leading to impaired force transmission and dilated cardiomyopathy [34] . Consistent with the arrhythmogenic lethal phenotype , HSPB7 CKO mice exhibited severe disruption of the ID structure with mislocalization of N-cadherin and desmoplakin in the cardiomyocyte cytosol . Likewise , we speculate that loss of HSPB7 may cause the structural instability of ID components and further affect the localization and function of N-cadherin and desmoplakin . Furthermore , several genetic mutations of desmoplakin [35] and plakoglobin [36] have been found to cause arrhythmogenic cardiomyopathy ( AC ) in humans . AC is a complex disorder and is considered to be a progressive disease of the IDs with clinical manifestations , including progressive loss of cardiomyocytes , inflammatory infiltrates , and compensatory replacement with fibrofatty tissue , leading to HF , severe ventricular tachyarrhythmias , and sudden cardiac death . HSPB7 CKO mice do not present with the typical phenotype seen in AC patients and lacks severe DCM and fibrofatty replacement phenotypes . It is possible that the mutant animals die too soon from sudden death ( within 2 weeks ) to observe a long-term compensation effect . Adherens junctions and desmosomes are organized independently of gap junctions [37] . The loss of protein in either the adherens junctions or desmosomes can decrease the expression of gap junction proteins [38] and results in arrhythmogenesis development [33 , 39] . The HSPB7 CKO mutant presents a much more severe phenotype compared with connexin 43 CKO and other adherens junctions or desmosome mutant mice [30 , 33 , 40] , suggesting loss of HSPB7 could affect more than one of these protein functions . These results suggest that HSPB7 has a pivotal role as an ID protein for maintaining the structure and functional complexes of IDs . The molecular mechanism by which HSPB7 maintains ID structure remains unclear . Our previous study identified FLNC , an actin-binding protein , as the HSPB7 interaction protein [28] . Loss of HSPB7 in skeletal muscle causes myofibrillar disorganization and sarcolemma disruption [28] . FLNC participates in the attachment of the sarcomere’s Z-lines to the costamere and IDs allowing cell-to-cell mechanical force transduction [41] . In the present study , myofibril organization appeared distorted in the mutant hearts that had wider , less dense Z-lines and loose actin filaments . Such a phenotype may have been caused by defects in the IDs . Ablation of ID proteins , such as N-cadherin [33] and plakoglobin [42] , can lead to distortion of the sarcomere , which may reflect loss of myofibril tension because of a lack of myofibril anchorage at the plasma membrane . Alternatively , loss of FLNC can also result in extensive disruption of the thick and thin filaments and the loss of distinct Z-lines in mice [43] . Furthermore , the down regulation of connexin 43 is before the occurrence of FLNC up-regulation and aggregation ( d4 ) ; and the mislocalization of desmoplakin or N-cadherin in HSPB7 CKO cells is not always detected with the aggregation of FLNC . Taken together , our results suggest that the up-regulation of FLNC may not be the first step in the resulting phenotype . Loss of HSPB7 would cause the structural instability of FLNC and then trigger the chaperone-assisted selective autophagy ( CASA ) pathway to increase the gene expression and aggregation of FLNC [28] . The CASA machinery can incorporate tension sensing , autophagosome formation , and transcription regulation to maintain filamin protein homeostasis in mammalian cells [44 , 45] . Loss of HSPB7 would facilitate FLNC unfolding or conformational changes that would further activate the CASA pathway , thereby leading to FLNC upregulation and then aggregation . A recent study reported that 23 different truncating variants in FLNC are highly associated with variable fractures of DCM and AC [46] . The presence of FLNC aggregates has also been identified in cardiomyocytes of patients with cardiomyopathy [47] and cardiac arrhythmia [48] . However , our results indicated that loss of HSPB7 resulting in the mislocalization of desmoplakin or N-cadherin would not be directly affected by the effect of FLNC overexpression and aggregation . On the other hand , the possibility still cannot be ruled out that the interaction between HSPB7 and FLNC is required for maintaining IDs and myofibrillar functional structures in the adult heart . In conclusion , our results provide the first comprehensive study characterizing HSPB7 as an ID protein and revealing information regarding the biological function of HSPB7 in the adult myocardium . The loss of HSPB7 results in the disruption of the ID structure with abnormal cardiac conduction function and thus induces arrhythmic sudden death , indicating the phenotype in the HSPB7 CKO mice is at least partly similar to that in human AC patients . Although it has not yet been reported , given the severity of the cardiac phenotype in our animal model , the functional mutation variants of HSPB7 gene could be identified in patients with AC . Thus , our mouse model demonstrates that HSPB7 is required for the structural integrity and function of gap-junction complexes and IDs , a finding which has vital implications for human heart disease .
All animal experiments were performed in accordance with the guidelines established by the Institutional Animal Care and Use Committee ( IACUC ) of Academia Sinica . All the experimental protocols were approved by IACUC and the approval number is 10–12–113 . Mice were treated according to institutional protocols or perfused transcardially under deep Avertin anesthesia for further perfusion . To generate inducible cardiac-specific HSPB7 CKO mutants , transgenic mice expressing a tamoxifen-inducible Cre recombinase protein under the control of the α-myosin heavy chain promoter , αMHC/MerCreMer mice ( MCM ) [26] , were intercrossed with HSPB7Flox/Flox mice [28] in a mixed 129S6/SvEvTac with a C57BL/6 genetic background . To induce Cre recombination , 8- to 10-week-old male MCM/HSPB7Flox/Flox mice were treated with 40 mg/kg of tamoxifen ( cat#L5647 , Sigma ) by intraperitoneal injection for 4 consecutive days . Tamoxifen was dissolved in corn oil at a concentration of 10 mg/mL heated to 37°C for 1 h . Mice were sacrificed at 4 and 7 days following the initiation of the tamoxifen treatment , and Cre-negative littermates were used as controls . All animal experiments were performed using protocols approved by the Institutional Animal Care and Use Committee , IBMS , Academia Sinica . For tissue extracts , the hearts were dissected from anesthetized mice , rinsed with cold PBS , blotted dry , weighed , and then homogenized using TissueLyser LT ( QIAGEN ) in 1X Cell Lysis Buffer ( cat#9803 , Cell Signaling Technology ) with a complete protease inhibitor cocktail ( cat#11836145001 , Roche Applied Sciences ) . For fractionation assay , mouse hearts were homogenized in lysis buffer containing 9M urea . The insoluble pellet fraction was sedimented through centrifugation at 16 , 000 g for 15 min . An equal volume of 2X SDSPAGE gel sample buffer was added to the supernatants and insoluble pellet fraction . After heating at 100°C for 10 min , the supernatants were stored at −80°C as aliquots until use . For western blotting experiments , protein extracts ( 30 μg of total protein ) were separated on 8% or 15% polyacrylamide gels and blotted onto PVDF membranes ( Millipore ) . Membranes were blocked with blocking buffer ( 5% nonfat dry milk , 10 mM of Tris–HCl , pH 7 . 6 , 150 mM NaCl , and 0 . 1% Tween 20 ) and incubated with the primary antibody at 4°C overnight . After incubation with peroxidase-conjugated secondary antibodies , proteins were visualized using enhanced chemiluminescence reagents ( Millipore ) and detected using an ImageQuant LAS 4000 mini system ( GE Healthcare Life Sciences ) . Densitometric analyses were performed using ImageJ software . The protein levels of GAPDH were used to normalize the results . The primary antibodies used included mouse monoclonal antibodies anti-α-actinin ( clone EA-53 , Sigma-Aldrich; 1:1000 . ) , anti-Actin , cardiac ( clone AC1-20 . 4 . 2 , Sigma-Aldrich; 1:1000 ) , anti-connexin43 ( clone CXN-6 , Sigma-Aldrich; 1:1000 ) , anti-GAPDH ( clone 6C5 , Millipore Corporation; 1:3000 ) ; and anti-Vinculin ( clone hVIN-1 , Sigma-Aldrich , Inc . ) ; rabbit polyclonal antibodies anti-FLAG ( F7425 , Sigma-Aldrich; 1:1000 ) , antipan-cadherin ( #C3678 , Sigma-Aldrich; 1:1000 ) ; goat polyclonal antibodies anti-FLNC ( K-18 , Santa Cruz Biotechnology; 1:200 ) ; and guinea pig polyclonal antibody anti-HSPB7 ( G11W , LTK BioLaboratories; 1:1000 ) . Mouse heart tissues were collected , fixed with 10% formalin , buffered with phosphate , and embedded in paraffin . Tissue sections ( 5 μm ) were subjected to hematoxylin and eosin and Masson’s trichrome staining using standard procedures [49] . For immunofluorescence staining , heart tissue was isolated from HSPB7 CKO mice , directly embedded in optimal cutting temperature compound ( OCT ) , and cryosections ( 16-μm sectioned ) were prepared . The sections were postfixed in 2% paraformaldehyde , blocked with 2% bovine serum albumin , and incubated with primary antibodies at 4°C overnight . The primary antibodies used included mouse monoclonal antibodies anti-α-actinin ( clone EA-53 , Sigma-Aldrich; 1:200 . ) , anti-Actin , cardiac ( clone AC1-20 . 4 . 2 , Sigma-Aldrich; 1:200 ) , anti-connexin43 ( clone CXN-6 , Sigma-Aldrich; 1:200 ) , antidesmoplakin ( clone DP2 . 15 , Millipore Corporation; 1:200 ) , anti-mMaC myomesin B4 ( DSHB; 1:200 ) , and anti-Vinculin ( clone hVIN-1 , Sigma-Aldrich ) ; rabbit polyclonal antibodies anti-FLAG ( F7425 , Sigma-Aldrich; 1:200 ) , anti-pan-cadherin ( #C3678 , Sigma-Aldrich; 1:200 ) ; goat polyclonal antibodies anti-FLNC ( K-18 , Santa Cruz Biotechnology; 1:100 ) ; and guinea pig polyclonal antibody anti-HSPB7 ( G11W , LTK BioLaboratories; 1:200 ) . After washing in PBS , sections were incubated with secondary antibodies , including FITC- or Rhodamine-conjugated goat anti-mouse and anti-rabbit IgG , FITC- conjugated donkey anti-goat IgG , Rhodamine-conjugated donkey anti-guinea pig IgG secondary antibodies ( Jackson ImmunoResearch Laboratories ) . Fibrosis was detected by staining using Alexa Fluor 647 wheat germ agglutinin ( WGA , Invitrogen; 10 μg/ml in PBS ) for 10 min at room temperature . Counterstaining was performed using 0 . 5 μg/ml of Hoechst 33342 ( Cell Signaling Technology ) . Fluorescence was visualized using a Zeiss LSM700 confocal microscope . Two-dimensional echocardiography was performed using a Philips iE33 ultrasound imaging system ( Philips Medical Systems , Best , Netherlands ) equipped with a 7–15 MHz linear array transducer on 10- to 12-week-old ( n = 5 per group ) mice . Echocardiography was recorded before tamoxifen administration and at 4 or 7 days after the first tamoxifen treatment for changes in cardiac function . The animals were initially anesthetized using 3% isoflurane . After the animals were sedated , anesthesia was maintained using 1% isoflurane during the echocardiographic examination . Heart rate was maintained between 350 and 600 beats per minute . After two-dimensional long- and short-axis images of the left ventricular ( LV ) were obtained , M-mode traces were acquired for measurement of the LV chamber dimensions at the diastole and systole , as well as the wall thickness . Echocardiography-derived LV mass , fractional shortening ( FS ) , and ejection fraction were recorded . Measurements were averaged from five consecutive cardiac cycles . Surface ECGs were recorded using a PowerLab 8/30 ( AD Instruments , Dunedin , New Zealand ) . ECGs were used to assess mice before and at 7 days after the first tamoxifen treatment . The animals were anesthetized using 1% isoflurane and standard three-lead surface ECG recordings were performed continuously for 15 minutes . The data were digitized and stored for off-line analysis using LabChart Pro software ( AD Instruments ) . For telemetric measurement of the ECG , miniature telemetry transmitter devices ( HD-X11 , Data Sciences International ) were implanted subcutaneously on the back with electrodes surgically placed and sutured to the right of the trachea and the left upper abdominal region as described [50] . The animals were allowed 7 days to recover from the surgery before telemetry recordings were acquired . Recordings from control and CKO mice were assessed continuously for 2 hours before and at 4 , 7 , and 14 days after the first tamoxifen treatment . For CKO mice , recordings were continuous beginning at day 7 after injection until the time of death . The receiver ( Physiotel Receiver , model RPC-1; Data Sciences International ) was used for data acquisition . ECG signals were digitized and stored for off-line analysis using LabChart Pro software ( AD Instruments ) . Mouse left ventricle tissue was diced into small blocks in a fixative mixture of glutaraldehyde ( 1 . 5% ) and paraformaldehyde ( 1 . 5% ) in phosphate buffer at pH 7 . 3 . The procedure was identical to that described previously [51] . Ultrathin sections were cut , mounted , post-stained , and observed using a FEI TECNAI G2 F20 S-TWIN electron microscope ( Electron Microscope Core Facility , Institute of Cellular and Organismic Biology , Academia Sinica ) . For immunoprecipitation experiments , hearts from adult C57BL/6 wild-type and HSPB7Flox/Flox mice were lysed in Cell Lysis Buffer ( Cell Signaling Technology ) with complete protease inhibitor cocktail ( Roche Applied Sciences ) . Subsequently , 1 mg of total protein extracts were incubated with 50 μL of prewashed anti-FLAG M2 affinity gel ( A2220 , Sigma-Aldrich ) at 4°C overnight . The beads were then washed with PBS and analyzed through western blotting with an anti-FLNC antibody ( K-18 , Santa Cruz Biotechnology ) . Wild-type mice heart was used as a negative control . Anti-HSPB7 ( G11W , LTK BioLaboratories ) was used as a control for the equal loading of heart extracts . For in vivo tests of muscle cell membrane integrity , 8- to 10-week-old control and CKO mice ( n = 4 per group ) were first treated using 40 mg/kg of tamoxifen ( Sigma ) through intraperitoneal injection for four consecutive days . Evans Blue dye ( EBD; Sigma ) with 0 . 1 mg/g of body weight was intraperitoneally injected for two consecutive days into HSPB7 CKO and control mice 5 days after the first tamoxifen administration . After 18 h , mice were sacrificed and their hearts were harvested and cryosectioned . EBD-positive myofibers were directly observed under a stereomicroscope with blue color ( SMZ1500 , Nikon ) and a fluorescence microscope with red autofluorescence ( BX51 , Olympus ) . The adenoviruses Ad-Cre ( 5 × 107 pfu per μL ) for Cre recombinase were kindly provided by Guey-Shin Wang ( Institute of Biomedical Sciences , Academia Sinica , Taiwan ) . To achieve successful gene delivery through intracardiac virus infection , 10- to 12-week-old male HSPB7Flox/Flox mice weighing 25–30 g were used in the experiments . Mice were intraperitoneally anesthetized with Avertin at a dose of 100 mg/kg . The skin was incised at the level of the left third and fourth ribs and the pectoral muscles were dissected using two fine forceps and retracted gently with a 6–0 silk suture to free the location . The 27-g needle was inserted 4-mm deep directly into the thorax between the third and fourth ribs . Then , we performed four injections ( 10 μL each ) of Ad-Cre ( 5 × 107 pfu/μL ) into the cardiac wall . After a slow injection , the pectoral muscles were quickly released and the skin was sutured using a 3–0 silk suture , and animals were observed and monitored until recovery . After sacrificing the mice at d8 , heart tissue was isolated , directly embedded in OCT , and cut ( cryosectioned , 16-μm sections ) for further immunofluorescence study . Results are presented as mean ± s . d . Comparisons between the two groups employed two-tailed Student's t-test . Mouse survival rates were calculated through the Kaplan–Meier method . When analyzing statistical differences between the groups of mice , a P value of less than 0 . 05 was considered significant .
|
The intercalated disc is an indispensable structure that connects neighboring cardiomyocytes . It is also considered to be a single functional unit for cellular electric , mechanical , and signaling communication to maintain cardiomyocyte rigidity and synchrony . Mutation or defect in intercalated disc components usually results in distortions in the structure of intercalated discs and lethal cardiac abnormalities in patients . In this study , we found that the dynamic expression and subcellular location of HSPB7 are highly associated with intercalated disc component protein , N-cadherin , during the assembly and maturation of intercalated discs in cardiomyocytes . To identify the functional role of HSPB7 in the adult heart , we conducted a loss-of-function study of HSPB7 using a gene conditional knockout approach . We found that the loss of HSPB7 quickly results in the disruption of the intercalated disc structure , decreasing the expression of connexin 43 and mislocalization of N-cadherin and desmoplakin , and further inducing arrhythmic sudden death . In conclusion , our mouse model demonstrates that HSPB7 is required to maintain the structure and function of gap-junction complexes and intercalated discs , which has important implications for human heart disease .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"cell",
"physiology",
"medicine",
"and",
"health",
"sciences",
"myofibrils",
"muscle",
"tissue",
"cardiovascular",
"anatomy",
"nervous",
"system",
"molecular",
"probe",
"techniques",
"junctional",
"complexes",
"electrophysiology",
"cloning",
"neuroscience",
"animal",
"models",
"gap",
"junctions",
"electrocardiography",
"model",
"organisms",
"experimental",
"organism",
"systems",
"bioassays",
"and",
"physiological",
"analysis",
"molecular",
"biology",
"techniques",
"cardiac",
"muscles",
"research",
"and",
"analysis",
"methods",
"musculoskeletal",
"system",
"immunoblot",
"analysis",
"animal",
"cells",
"biological",
"tissue",
"mouse",
"models",
"muscles",
"electrophysiological",
"techniques",
"molecular",
"biology",
"muscle",
"cells",
"sarcomeres",
"cardiac",
"electrophysiology",
"anatomy",
"cell",
"biology",
"synapses",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"neurophysiology",
"heart"
] |
2017
|
HSPB7 prevents cardiac conduction system defect through maintaining intercalated disc integrity
|
Lifelong persistence of Epstein-Barr virus ( EBV ) in infected hosts is mainly owed to the virus' pronounced abilities to evade immune responses of its human host . Active immune evasion mechanisms reduce the immunogenicity of infected cells and are known to be of major importance during lytic infection . The EBV genes BCRF1 and BNLF2a encode the viral homologue of IL-10 ( vIL-10 ) and an inhibitor of the transporter associated with antigen processing ( TAP ) , respectively . Both are known immunoevasins in EBV's lytic phase . Here we describe that BCRF1 and BNLF2a are functionally expressed instantly upon infection of primary B cells . Using EBV mutants deficient in BCRF1 and BNLF2a , we show that both factors contribute to evading EBV-specific immune responses during the earliest phase of infection . vIL-10 impairs NK cell mediated killing of infected B cells , interferes with CD4+ T-cell activity , and modulates cytokine responses , while BNLF2a reduces antigen presentation and recognition of newly infected cells by EBV-specific CD8+ T cells . Together , both factors significantly diminish the immunogenicity of EBV-infected cells during the initial , pre-latent phase of infection and may improve the establishment of a latent EBV infection in vivo .
Epstein-Barr virus ( EBV ) is a ubiquitous human herpes virus with strong tropism for human B cells . EBV persists in an infected host for life by establishing a latent infection in B cells that represent an immunologically silent reservoir . Eventual reactivation of these cells into the lytic cycle leads to the production of progeny viruses that spread to other cells and hosts . The lytic phase goes along with the expression of high amounts of viral antigen , including the highly immunogenic immediate early proteins BZLF1 , BRLF1 and BMRF1 [1] . To protect lytically activated cells from immune recognition , EBV takes several measures to perturb anti-viral immune responses of the host ( [2] for review ) . EBV also expresses a set of immunogenic latent and lytic proteins immediately following the infection of target cells [3] , [4] , suggesting that also newly infected cells are prone to immune recognition . Hypothetically , the virus copes with this immunological challenge by active immune evasion in newly infected cells , which would be in close analogy to cells in the productive lytic phase . EBV codes for a number of proteins that subvert the host's immune surveillance ( [5] for review ) : a homologue of human IL-10 with anti-inflammatory properties ( vIL10 ) [6]–[8] , encoded by the EBV gene BCRF1 [9] , the DNAse/exonuclease BGLF5 that shuts off host protein synthesis [10] and contributes to Toll-like receptor 9 downregulation in productively infected cells [11] , the G-protein-coupled receptor BILF1 that degrades MHC class I molecules [12] , and BNLF2a , a protein unique to lymphocryptoviruses that inhibits the transporter associated with antigen processing ( TAP ) [13] . These proteins have so far been classified as lytic proteins and , correspondingly , were functionally investigated in lytic subsets of EBV-infected cell lines in vitro . Several viruses including EBV , primate cytomegaloviruses ( CMVs ) , Orf poxvirus , and equine herpes virus type 2 ( EHV-2 ) encode homologues of human IL-10 [9] , [14] , [15] , strongly suggesting that IL-10 is advantageous for these viruses . Accordingly , different immunomodulatory activities in infected and bystander cells have been described for viral IL-10 homologues including inhibition of DC maturation [16] and inhibition of Th1 cytokine expression [17] . Recently , it has been reported that vIL-10 of CMV has a profound impact on innate and adaptive immune responses in an in vivo model [18] EBV's IL-10 homologue has been described to be critical for B-cell growth transformation [6] , to block gamma interferon release [19] , to reduce MHC-I levels on B cells [8] and to functionally inhibit T cells [20] and monocytes [7] . vIL-10 encoded by BCRF1 is expressed early upon infection [19] , but its precise role and immunomodulatory capacities especially during the initial phase of EBV infection remain elusive . BNLF2a , in contrast , is unique to the family of lymphocryptoviruses , but many other viruses pursue analogous strategies of TAP inhibition ( [21] for review ) . BNLF2a prevents binding of both ATP and peptide to TAP and thereby prevents peptide loading to MHC class I molecules [22] . Ectopic expression of BNLF2a leads to reduced surface levels of MHC class I molecules [13] that are unstable without properly loaded peptides [23] . BNLF2a is expressed early in the productive lytic phase and reduces the recognition of B cells by T lymphocytes specific for viral immediate early and early lytic proteins [24] . In this study , we extend our knowledge about BNLF2a and vIL-10/BCRF1 . We show that both proteins contribute to the immune evasion of EBV in newly infected primary B cells . Both proteins are expressed immediately following infection . With EBV mutants deficient in BCRF1 and BNLF2a , we demonstrate that BNLF2a impairs the recognition of virally infected B cells by EBV-specific CD8+ T lymphocytes during the very first days of infection . Additionally , we identified vIL-10 to protect EBV-infected B cells from NK cell-mediated elimination . Furthermore , vIL10 released from newly infected B cells prevents the secretion of anti-viral cytokines , thereby abrogating anti-viral CD4+ effector T cell functions . In summary , BNLF2a and vIL-10/BCRF1 act in a complementary manner to prevent immune recognition and elimination of newly EBV-infected B cells .
Maxi-EBV genomes deficient in BCRF1 and/or BNLF2a were constructed by targeted mutation of the maxi-EBV plasmid p2089 [25] . Maxi-EBV mutagenesis was performed by homologous recombination in accordance to previous work [26] . BCRF1 deletion mutants were generated by replacing the entire gene by a prokaryotic kanamycin resistance expression cassette . The BNLF2a locus of EBV is complex ( Figure S1A ) . BNLF2a shares its transcript with BNLF2b and is situated in the first intron of the TP gene encoding the latent membrane protein ( LMP ) 2A . Moreover , this genomic locus is part of the 3′ untranslated region of BNLF1 encoding LMP1 . To abrogate BNLF2a expression , the first translational start codon of BNLF2a was mutated to a stop codon preventing BNLF2a translation . The exchange of only four nucleotides reduced the risk of interfering with the expression or regulation of adjacent genes . In total , we constructed three EBV mutants: two single mutants that were null for BCRF1 ( ΔBCRF1 ) or BNLF2a ( ΔBNLF2a ) and one double mutant that combined both functional deletions ( double k . o . ) . Technical details , cloning strategies , and restriction enzyme digests confirming BAC integrity are provided in Material & Methods and Figure S1 . We established single cell clones from HEK293 cells stably transfected with the mutant viruses described above by selecting for hygromycin resistance . Clonal cells lines were tested for virus production upon transfection of an expression plasmid encoding the lytic transactivator BZLF1 [27] . The titers of infectious virus in the supernatants of these clones were calculated as described in Material & Methods . The genotypes of selected clones were confirmed by Southern blot hybridization ( Figure 1A ) and infected B cells were routinely tested by PCR to confirm infection with the respective virus mutant ( Figure 1B ) . EBV expresses a set of lytic genes during the initial , pre-latent phase of B-cell infection [3] , [4] and EBV virions contain a variety of viral RNAs [28] , which prompted us to address the expression kinetics of the immunomodulatory proteins vIL-10/BCRF1 and BNLF2a during pre-latent infection . For this , we infected primary peripheral B cells with 2089 wild-type EBV or with the ΔBNLF2a , ΔBCRF1 or double k . o . mutant viruses . We then prepared cDNA from infected cells at different time points post infection ( p . i . ) and assessed the expression of the BCRF1 gene as well as levels of the bicistronic transcript encoding BNLF2a and BNLF2b by quantitative RT-PCR ( qPCR ) . Figure 2A shows that both transcripts were detectably present as early as one day p . i . The comparison to glucuronidase beta ( GUSB ) transcripts , a validated housekeeping gene in LCLs [29] , revealed that BNLF2a/b expression levels increased initially , followed by a plateau , whereas BCRF1 transcript levels declined during the first days p . i . before reaching a stable level . Performing flow cytometry , we could demonstrate the rapid expression of BNLF2a protein in cells infected with 2089 wild-type EBV and ΔBCRF1 mutant EBV , but not in cells infected with ΔBNLF2a or double k . o . mutant viruses ( Figure 2B ) confirming the genetic ablation of BNLF2a . No specific vIL-10-antibody was available to confirm BCRF1 deficiency . BNLF2a interferes with antigen presentation on MHC class I molecules by inhibiting the transporter associated with antigen processing ( TAP ) [13] , [22] . EBV-specific CD8+ T-cell clones constitute sensitive tools to measure antigen presentation of EBV-infected B cells in vitro . In order to analyze BNLF2a effects during the earliest phase of infection , we infected primary peripheral B cells with 2089 wild-type EBV or with the mutant viruses ΔBCRF1 , ΔBNLF2a , or double k . o . and used them as targets for clonal CD8+ T cells . One of these T-cell clones detects the HLA-B8-restricted epitope RAKFKQLL ( RAK ) derived from BZLF1 protein [30] , the master regulator of the lytic cycle [31] . Co-cultures at defined effector/target ratios were prepared at different days after B-cell infection and incubated overnight . ELISA assays on gamma interferon ( IFN-γ ) levels in the supernatant were indicative of T-cell activation . The experiments revealed that RAK-specific T cells recognized B cells infected with either ΔBNLF2a or the double k . o . mutant virus significantly better than B cells infected with either 2089 wild-type or ΔBCRF1 mutant EBVs ( Figure 3A ) . Of note , the difference in recognition became evident already on day 1 p . i . At this time point , only B cells infected with ΔBNLF2a or double k . o . mutant viruses detectably activated the RAK-specific T cell clone , whereas cells infected with 2089 EBV or the ΔBCRF1 mutant virus did not . The level of recognition of infected B cells peaked on day 4 p . i . and then declined until it reached similar levels as in control LCLs , indicating the establishment of latency . Similar results were obtained with CD8+ T-cell clones specific for the epitopes QAKWRLQTL ( QAK ) ( Figure 3B ) and IEDPPFNSL ( IED ) ( Figure 3C ) derived from the latent proteins EBNA3a and LMP2a , respectively , further emphasizing the immunoevasive function of BNLF2a in freshly infected cells . In contrast , the response of clonal CD8+ T cells specific for the CLG epitope of LMP2 protein was independent of BNLF2a ( Figure 3D ) . This epitope is known to be loaded TAP-independently onto MHC I molecules because its high hydrophobicity presumably allows for passive diffusion through the ER membrane [32] . The delay in T-cell recognition as compared to gene expression is presumably attributable to the fact that EBV-infected B cells reach their full antigen presenting potential not until a few days p . i . [33] or to antigen immunodominance [34] . To ensure that our genetic manipulation did not alter expression levels of the investigated antigens , we performed quantitative PCR that revealed similar expression levels in B cells infected with the different viruses ( Figure S2 ) . The previous experiments revealed that vIL-10 did not influence the recognition of freshly infected B cells by EBV-specific CD8+ T cell clones in vitro ( Figure 3 ) . Nevertheless , its conservation in different EBV isolates [35] and the strong immunomodulatory capacity of its cellular homologue [36] both suggest a prominent role for BCRF1/vIL-10 in EBV infection . IL-10 is known to sustainably promote Th2 cytokine responses [36] , which prompted us to assess the influence of vIL-10 on the secretion of various Th1/Th2 cytokines by PBMCs in response to an EBV infection . We prepared PBMCs from a donor with EBV immunity , determined the B cell content and infected them with 2089 EBV , ΔBCRF1 , ΔBNLF2a , or double k . o . mutant viruses with 0 . 1 GRU/B cell . We cultured these EBV-infected PBMCs for twelve days and evaluated the levels of Th1 and Th2 cytokines in the supernatants by multiplex ELISAs . The cytokine composition in the supernatants differed between PBMCs infected with the BCRF1-positive viruses ( 2089 EBV , ΔBNLF2a ) and mutant viruses lacking BCRF1 ( ΔBCRF1 , double k . o . ) . In detail , PBMCs infected with the ΔBCRF1 or double k . o . mutant viruses produced significantly higher levels of the pro-inflammatory cytokines IFN-γ , IL-2 , IL-6 , and TNF-β and of anti-inflammatory IL-10 ( Figure 4A ) , whereas levels of IL-1 , IL-5 , IL-8 , and TNF-α were not affected ( not shown ) . Interestingly , we observed the highest levels of the hIL-10 in the supernatants of PBMCs that were infected with either of the BCRF1-lacking mutant viruses ΔBCRF1 or double k . o . This observation points to the regulation of hIL-10 by vIL-10 or to increased IL-10 release in the course of a secondary cytokine response evoked by the high levels of pro-inflammatory cytokines . Besides activated T cells , NK cells can specifically lyse virus-infected cells . Regarding EBV , NK cells preferentially eliminate infected cells in lytic phase [37] . To address the question whether NK/NKT cell-mediated lysis differed between B cells infected with 2089 wild-type or mutant EBVs in the pre-latent phase , we infected purified peripheral B cells and added autologous purified CD56+ NK/NKT cells on day 3 p . i . We then assessed specific B-cell lysis after 3 hours of co-incubation and observed a significantly stronger lysis of B cells infected with the ΔBCRF1 or double k . o . mutant viruses as compared to B cells infected with 2089 EBV or the ΔBNLF2a mutant virus ( Figure 4B left panel ) . In a parallel experiment , we included CD4+ T cells that represent an important source for many cytokines . We hypothesized that CD4+ T cells could provide a supporting microenvironment for NK/NKT cells . Indeed , we found that the presence of CD4+ T increased NK/NKT-mediated lysis of infected B cells . This adjuvant effect was most pronounced when B cells were infected with either ΔBCRF1 or double k . o . mutant viruses ( Figure 4B mid panel ) . The difference became most evident at lower effector/helper/target ratios ( Figure 4C ) . In contrast , CD4+ T cells did not significantly increase NK/NKT-mediated lysis of B cells with 2089 wild-type or ΔBNLF2a mutant EBV ( Figure 4B , middle panel ) . CD4+ T cells alone did not reveal any cytolytic activity above background ( Figure 4B , right panel ) . The cytolytic activity of NK cells is regulated by MHC class I levels and by accessory molecules such as ligands of the activating natural killer group 2 member D ( NKG2DL ) , comprising MICA , MICB , and UL16-binding proteins ( ULBP ) 1–6 . Since IL-10 was described to modulate MHC surface levels [8] and expression of NKG2D ligands [38] , we investigated the expression of these molecules in more detail . Flow cytometry revealed that newly infected B cells displayed different levels of MHC class I at their surfaces varying in a temporal fashion throughout the first days of infection ( Figure S3A ) . However , this pattern of MHC I surface levels was not affected by the use of 2089 EBV or mutant virus . MHC I/BNLF2a double staining of GFP-positive , i . e . infected B cells 3 days p . i . particularly correlated that endogenous expression levels of BNLF2a did not alter the surface levels of MHC class I molecules of cells infected with different viruses ( Figure S3B and C ) . Hence , improved NK killing of B cells infected with either ΔBCRF1 or double k . o . mutant EBVs was not attributable to different MHC class I levels . Next , we analyzed the expression levels of members of the family of NKG2D ligands by qPCR . With the exception of ULBP4 and 6 , all ligands were clearly expressed . Importantly , expression kinetics were independent of the virus mutant used for infection ( Figure S4 ) . As NK and CD4+ T cells express the IL-10R ( [39] and Figure S5 ) our data suggested a direct effect of vIL-10 on these effector cells . This hypothesis was further substantiated by rescue experiments: the addition of exogenous viral and human IL-10 at physiological concentrations ( 1 ng/ml ) partially reverted NK-mediated killing of infected B cells , and exogenous vIL-10 completely reverted CD4+ T cell assistance ( Figure S6 ) . K562 cells do not express MHC I molecules and are therefore efficiently killed by NK/NKT cells . Intriguingly , the same concentration of IL-10 that inhibited killing of newly EBV-infected B cells did not affect NK/NKT-mediated killing of K562 cells ( Figure S7 ) . Taken together , our results indicate that vIL-10 directly impairs NK/NKT-mediated lysis of newly EBV-infected B cells and inhibits CD4+ T cell support of NK-mediated killing . Regression assays are a means to quantify EBV-specific memory T-cell responses in vitro . In such assays , experimentally infected PBMCs from EBV-positive donors show regression of B-cell outgrowth in vitro , reflecting the reactivation of an EBV-specific memory T-cell response [40] . The strength of regression depends on the number of EBV-specific immune effectors , which mirror the extent of the donor's EBV immunity as well as the ability of infected cells to escape immune elimination . Identical EBV-mediated transformation of primary B cells is a prerequisite when comparing different viruses in regression assays . Therefore , we initially determined the dose-dependent transformation of B cells by wild-type and mutant EBVs in limiting dilution assays . Purified peripheral B cells were infected with serial dilutions of the different virus stocks and the number of outgrowing lymphoblastoid cells was scored six weeks p . i . As shown in Figure 5A , we did not observe any differences in the rates of B-cell transformation . In a next series of experiments , we infected serial dilutions of PBMC preparations from EBV-seropositive donors with 2089 wild-type EBV or the ΔBCRF1 , ΔBNLF2a , or double k . o . mutant viruses and seeded the cells in 96-well cluster plates . Six weeks after infection we analyzed cell viability in an MTT assay . B cells infected with ΔBCRF1 , ΔBNLF2a , or wild-type EBV were killed equally but B cells infected with the double k . o . mutant EBV were eradicated much more efficiently ( Figure 5B ) . Thus , deletion of vIL-10 and BNLF2a synergistically affected the outgrowth of infected B cells in the presence of EBV-specific immune effectors in vitro . This finding strongly suggests that pre-latent vIL-10- and BNLF2a-mediated immune evasion contributes to the success of EBV infection also in its native host . In line with results published by others [41]–[43] , subsequent analyses with CD4-depleted PBMCs indicated that CD4+ T cells are essential for regression of EBV-infected B cells in vitro ( Figure 5C ) . In this setting , immune control of B cells infected with 2089 wild-type or ΔBNLF2a mutant EBVs was completely abrogated , while regression of B cells infected with either ΔBCRF1 or double k . o . mutant virus was partially maintained . These results suggested that helper CD4+ T cells contribute to regression , probably by providing stimulatory cytokines such as IL-2 to CD8+ effector T cells [41] , and vIL-10 directly interferes with reactivation of EBV-specific memory CD8+ T-cells . Consistent with published data [41] , the contribution of CD56+ NK/NKT cells to the regression of EBV-infected PBMCs was less pronounced , but immune control of cells infected with 2089 wild-type or ΔBNLF2a mutant EBVs was slightly reduced in comparison to non-depleted PBMCs , which is again in line with a potential inhibitory effect of vIL-10 on CD4+ or CD8+ T cells ( Figure 5D ) .
EBV persists for life in its host despite the presence of strong anti-viral immune responses . The asymptomatic co-existence depends on an immunological equilibrium of anti-viral activities of the host and viral counter-mechanisms . Thus , reduced viral protein expression during latency as well as active immune evasion is essential for EBV . Recently , a number of viral strategies of active immune evasion during the lytic phase have been identified for EBV ( [5] for review ) . Especially immediate-early and early lytic proteins are among the most immunodominant EBV antigens [44] rendering their expression a particular immunological challenge for the virus . Recent reports described that immediate-early and early lytic genes are also expressed in newly infected cells following infection [3] , [4] putting these cells at risk for rapid elimination by immune effectors . Current data from our lab revealed that EBV virions deliver viral mRNAs , including those encoding the immunoevasins vIL-10 and BNLF2a , into target cells where they are instantly translated [28] . Along this line , our findings demonstrate that specific CD8+ T cells can recognize EBV infection of B cells as soon as one day p . i . Our observation that the RAK epitope of BZLF1 is presented instantaneously after infection fits the data of others on early BZLF1 expression in newly infected cells [3] , [4] . The immediacy of its expression may contribute to the observed immunodominance of BZLF1 among CD8+ target antigens of EBV [34] . This immunodominance may be shaped by cross-competition between CD8+ T cells for antigen-presenting cells , a process that favors T-cell responses against the earliest antigens presented during the process of infection [45] . EBV expresses the immune modulators vIL-10 and BNLF2a as early as six hours p . i . and actively perturbs the host's immune response to newly infected cells . These findings further indicate that the pre-latent phase of EBV is critical for the outcome and success of viral infection . Of note , our T-cell experiments confirm that BNLF2a blunts activation of EBV-specific CD8+ T cells as early as one day p . i . As BNLF2a was reported to interfere with MHC peptide loading , destabilized MHC results in reduced surface levels that might render the infected cells vulnerable towards NK cells . However , B cells infected with 2089 wild-type EBV or ΔBNLF2a mutant virus were equally lysed by NK/NKT cells ( Figure 3B ) and displayed similar MHC class I surface levels ( Figure S3 ) . Hence , expression of BNLF2a seems to be tightly balanced in that it impairs loading of new antigenic peptides without reducing MHC-I surface levels during the pre-latent phase of infection . Expression of vIL-10 did not impair recognition of EBV-infected cells by clonal , EBV-specific CD8+ T cells ( Figure 2 ) . Cultured T-cell clones are pre-activated and the endogenous expression level of vIL-10 might be insufficient to repress these effectors . However , we detected direct effects of vIL-10 on ex vivo-isolated NK/NKT cells as well as CD4+ T cells . NK cells lyse EBV-infected B cells preferentially when they enter the productive lytic cycle [34] . We demonstrate here that NK cells also lyse newly infected B cells but vIL-10 interferes with this effector function ( Figure 4 ) . The presence of CD4+ T cells further supported NK-mediated target lysis , especially when vIL-10 was not expressed . This phenomenon is probably attributable to two different observations: ( i ) infections with BCRF1-deficient viruses led to higher levels of Th1 cytokines ( Figure 4A ) suggesting that increased Th1 cytokine secretion by CD4+ T cells boosted NK cell activity , and ( ii ) rescue experiments indicated a direct inhibitory effect of vIL-10 , as well as hIL-10 , on NK and CD4+ T cell activity ( Figure 4B ) . Our observation that BNLF2a and vIL-10 reduce the recognition of pre-latently infected B cells in the absence of MHC class I downregulation is in apparent contrast to earlier results pointing to downregulation of MHC class I by BNLF2a [13] or by vIL-10 [8] . It appears that BNLF2a can strongly decrease total MHC class I levels when ectopically expressed at high levels [13] , weakly so when endogenously expressed in lytically EBV-infected cells [24] , and not to a detectable extent when transiently expressed in the pre-latent phase ( our present data ) . In lytically EBV-infected lymphoblastoid cells , the presentation of relevant EBV epitopes , including the BZLF1 RAK epitope , was reduced by BNLF2a much more strongly than were total MHC class I levels [24] . Thus , our observations that overall MHC class I levels are maintained in pre-latently infected cells , while T cell recognition is reduced , can be well reconciled . Different considerations apply to the effects of vIL-10 on MHC class I levels . In our earlier study [8] , we showed that exogenous soluble vIL-10 or human IL-10 , as well as supernatants from the EBV-producing simian B cell line B95-8 , reduced total MHC class I levels on primary human B cells after 2 days' incubation . EBV-containing B95-8 supernatant contains significant amounts of IL-10 ( own observations ) . The conditions carried out with B95-8 supernatants might therefore mirror biological situations in secondary lymphoid organs in the amplification phase of infectious mononucleosis [46] , [47] . Whether such high IL-10 concentrations are still beneficial for the onset of infection or rather trigger NK cell activity is however questionable . In contrast , the conditions in our present study ( 12 h incubation of primary B cells in infectious supernatant and subsequent exchange of media ) potentially correspond to the conditions during acquisition of EBV from another virus carrier or spread of the virus for infection maintenance in the context of functional immune control . Our data demonstrate that pre-latent expression levels of endogenous viral IL-10 and BNLF2a together are balanced to avoid the reduction of MHC I surface levels in the pre-latent phase . Non-maximal expression necessarily results in more subtle effects , but complementarity of vIL-10 and BNLF2a might compensate for that . Hence , together both factors succeed to blunt cellular immune responses: while pre-latently expressed vIL-10 appears to act mainly on NK and T helper cells in a paracrine fashion , BNLF2a specifically blocks antigen presentation on MHC class I in infected cells . The overall effects of the two immunoevasins BNLF2a and vIL-10 on establishment of EBV infection in a complex immunological environment were studied in regression assays . In this experimental setting , we observed a phenotype with the double k . o . mutant virus , only , revealing a synergistic effect between the two immunoevasins ( Figure 5A ) . As viral functions within the first days of infection are decisive for the outcome of regression assays [48] , our result further emphasizes the importance of BNLF2a and vIL-10 . The experiments shown in Figure 5 also stress the role of CD4+ T cells in anti-viral immune responses , presumably by providing help to CD8+ cells [41] . Depletion of CD56+ NK cells had a minor effect , only , in accordance with previous observations [41] . This result and the observation that single ΔBNLF2a and ΔBCRF1 mutant viruses were as well controlled as wild-type EBV points at a degree of redundancy in immunological mechanisms of EBV control that can only be overcome by the combined action of viral immunoevasins with different mechanisms of action . Collectively , we demonstrate in this study that the immunoevasins vIL-10 and BNLF2a of EBV have important functions in B cells in the pre-latent phase immediately following infection . Together , these proteins interfere both with innate and adaptive immune responses and thus contribute to efficient immune evasion of newly infected B cells . These findings highlight that the pre-latent phase of EBV infection is decisive for successful establishment and persistence of the virus in its host .
The study was approved by the Ethics Committee of the Ludwig-Maximilians-Universität . Study participants or their legal guardians provided written informed consent . The mutant viruses generated for this work are based on the previously described 2089 EBV [25] . The strategies to replace BCRF1 and to block BNLF2a translation are depicted in Figure S1 . The maxi-EBV genomes were modified following an improved protocol for BAC recombineering in the E . coli strain SW105 [26] . The BCRF1 gene was replaced by a prokaryotic expression cassette for neomycin phosphotransferase II , conferring kanamycin resistance , flanked by homologous sequences ( 200 bp ) to the neighboring region of the gene . The DNA fragment was cloned , linearized and electroporated into recombination competent SW105 bacteria carrying the 2089 EBV genome . Transformants were selected for kanamycin resistance ( 50 µg/ml ) . The first codon ( Met1 ) of BNLF2a was replaced by a stop codon and concomitant insertion of an analytic SpeI site . First , a prokaryotic expression cassette for the galK gene flanked by 50 bp of homologous sequence to the nucleotides up- and downstream of the BNLF2a Met-1 codon was generated by PCR . The product was inserted into the 2089 EBV BAC by homologous recombination and SW105 clones were selected for competence in galactose metabolism . Then , a DNA fragment of 132 bp was synthesized comprising the four mutated nucleotides flanked by 64 bp of sequences homologous to the neighboring regions of the previously inserted galK cassette . Successfully modified clones were counter-selected for competence in galactose metabolism by growth on minimal plates with 2-deoxy-galactose ( DOG ) and glycerol as carbon sources . The genome of the double k . o . mutant virus ( deficient for BCRF1 and BNLF2a ) was generated by conversion of the BNLF2a-Met1 to a STOP codon in the ΔBCRF1 EBV genome . BAC integrity and the presence of the knockout-specific restriction sites were confirmed by sequencing regions of approx . 5 , 000 bp around the mutation sites . Primer sequences used for cloning are given in Table S1 . Recombinant EBV BAC-DNAs were prepared from bacteria , purified on a CsCl gradient and transfected into 293HEK cells using polyethylenimine ( Sigma-Aldrich , Munich , Germany ) . Cells were selected for hygromycin resistance ( 80 µg/ml , Life Technologies , Darmstadt , Germany ) and single clones were tested for GFP fluorescence . Virus production was induced by co-transfection of the expression plasmids p509 , encoding BZLF1 [27] , and p2670 , encoding BALF4 [49] . Supernatants were harvested three days later and cells and debris were removed by centrifugation and filtration through PVDF membranes ( 0 . 8 µm pore size ) . Titers of virus stocks ( referred to as ‘green Raji units’ , GRU ) were calculated as described [50] by infecting Raji B cells and measuring the number of GFP+ cells by flow cytometry three days later . Calculated titers were confirmed by infecting Raji cells with equal amounts of GRU from the different virus stocks resulting in equal percentages of GFP+ cells three days p . i . in all samples ( Figure S8 ) . Genomic DNA from producer clones was extracted , digested with restriction enzymes and analyzed by Southern blot as described [25] . Blots were hybridized against PCR-amplified DNA fragments derived from EBV's origin of replication , oriP , from the BNLF2b locus representing sequences adjacent to the BCRF1 and BNLF2a genes , respectively . Primer sequences are provided in Supplementary Table S2 . Primary B cells were either isolated from PBMCs of voluntary blood donors or buffy coats with the B cell isolation kit II ( Miltenyi ) yielding B cell populations of ≥95% purity . B cells were infected with EBV mutants at a multiplicity of infection ( MOI ) of 0 . 1 GRU/B cell and infected B cells were analyzed for their EBV genotype by PCR amplification of BCRF1 , the BNLF2a wild-type sequence or the BNLF2a k . o . ( Met1STOP ) sequence . The infection with recombinant EBV was confirmed with a PCR spanning GFP and a part of the 2089 EBV backbone . Primer sequences are provided online in Supplementary Table S2 . The recognition of EBV-infected B cells by EBV-specific CD8+ T cell clones was analyzed by IFN-γ ELISA assays . CD8+ specific T cell clones were established from PBMCs as previously described [51] . EBV-specificity of clonal CD8+ cells was verified by flow cytometry after staining with the respective TCR specific multimer ( Proimmune , Oxford , UK ) and a CD8-specific antibody ( BectonDickinson , Heidelberg , Germany ) . For recognition assays , triplicates of 10 , 000 specific CD8+ T cells and 20 , 000 HLA-matched infected B cells were co-incubated for 18 hours on 96 well cluster plates in a total volume of 200 µl . IFN-γ levels in the supernatants were measured by ELISA ( Mabtech , Nacka Strand , Sweden ) . Sample values were normalized to the IFN-γ level obtained from T cells co-incubated with an established EBV+ lymphoblastoid cell line ( LCL ) . The normalization corrected for putative changes in the performance of the T cells during the time course and inaccuracies in T cell counting . Cytokines in the supernatants of EBV-infected PBMCs were measured with the Th1/Th2 11-plex FlowCytomix Kit ( BenderMed Systems , Wien , Austria ) according to the manufacturer's instructions . Cells were washed in PBS , counted and stained with fluorophore-coupled antibodies against human CD3 , CD210 ( IL10R ) , CD56 , CD4 and MHC class I ( anti-human HLA A , B , C clone W6/32 ) ( Biolegend , Netherlands ) . For intracellular flow cytometry , cells were fixed and permeabilized using the Cytofix/Cytoperm Kit ( BD , Germany ) and blocked with FCS and TruStain FcX ( Biolegend ) prior to staining with BNLF2a-specific antibody MVH-8E2 ( a kind gift from A . Rickinson , Birmingham ) or isotype control . Cells were counterstained with APC-coupled anti-rat IgG F ( ab′ ) 2 fragments ( Jackson , Newmarket , UK ) . B cells isolated from PBMCs were infected with EBV mutant viruses , cultured for three days and then labeled with calcein ( Calcein AM , Life Technologies ) as described previously [52] . Autologous CD56+ cells and CD4+ cells were isolated by MACS sorting ( Miltenyi , Bergisch-Gladbach , Germany ) , analyzed by flow cytometry and used for further experiments in case of ≥95% purity . Infected B cells were then co-incubated with CD56+ and/or CD4+ cells in 96-well V-bottomed microtest plates at the indicated effector : target ratios with 1 unit representing 1 , 000 cells in a total volume of 200 µl . Three hours later , fluorescence in the supernatant was measured with a Wallac Victor plate reader ( Perkin-Elmer , Waltham MA , USA ) . Spontaneous calcein release of labeled cells without effectors was subtracted from sample values . Specific lysis represents the ratio of sample values to total lysis values . Total lysis was obtained by adding 1% Triton-X-100 to target cells . The transformation capacities of the recombinant viruses were assessed by limiting dilution assays as described [53] . In brief , serial dilutions of EBV mutant viruses were added to 48 replicates of 1×105 B cells prepared from adenoids in 96-well flat bottom plates and incubated for six weeks with weekly supply of fresh culture medium . Living cells were assessed by MTT assays [54] . Regression assays were performed as described elsewhere [48] . Complete PBMCs or PBMCs depleted of CD56+ or CD4+ cells by MACS sorting were infected overnight with EBV at a MOI of 0 . 1 GRU/B cell . Cells were supplied with fresh culture medium and seeded in 24 replicates on 96-well flat bottom microtest plates in serial dilutions , ranging from 100 , 000 to 100 cells/well . The number of wells with proliferating cells was assessed by MTT assays after six weeks of culture . B cells from adenoids were infected with EBV mutant viruses at a MOI of 0 . 1 GRU . Cells were harvested for total RNA preparation at different time points using the RNeasy MiniKit ( Qiagen , Hilden , Germany ) . Residual genomic DNA was removed by DNAse digestion and RNA was reversely transcribed with the SuperMix Kit ( Life Technologies ) . Quantitative PCR was performed on a LightCycler 480 ( Roche , Basel , Switzerland ) , using the SybrGreen LC480 Mix . Primer sequences are provided online in supplementary Table S3 .
|
Despite strong cellular and humoral immune responses , herpesviruses persist in their hosts for a lifetime . Epstein-Barr virus ( EBV ) is a herpesvirus that infects human B cells . This results in a latent infection where only a minimal set of viral proteins is expressed and infected cells cannot be eradicated by immune cells . When the virus reactivates in order to produce progeny , many viral proteins are expressed that are potential targets of immunity , but the virus coexpresses viral “immunoevasins” that blunt immune responses . Similarly , in the very first phase of B cell infection by EBV , called the pre-latent phase , a rather wide spectrum of antigens is expressed . However , it has been unknown whether viral immunoevasion occurs in this phase . Here we show that two viral immunoevasins are active in the pre-latent phase and prevent immune recognition by a variety of mechanisms: they reduce the presentation of EBV antigens to CD8+ killer T cells , prevent an attack by natural killer cells , and reduce the function of CD4+ helper T cells . Thus , it seems to be important for the virus to shield itself from attack by immune cells during the pre-latent stage .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"oncology",
"medicine",
"immune",
"evasion",
"basic",
"cancer",
"research",
"virology",
"biology",
"microbiology"
] |
2012
|
The EBV Immunoevasins vIL-10 and BNLF2a Protect Newly Infected B Cells from Immune Recognition and Elimination
|
Zika virus ( ZIKV ) is a mosquito-borne positive sense RNA virus . Recently , ZIKV emerged into the Western hemisphere as a human health threat , with severe disease associated with developmental and neurological complications . The structural envelope protein of ZIKV and other neurotropic flaviviruses contains an extended CD-loop relative to non-neurotropic flaviviruses , and has been shown to augment ZIKV stability and pathogenesis . Here we show that shortening the CD-loop in ZIKV attenuates the virus in mice , by reducing the ability to invade and replicate in the central nervous system . The CD-loop mutation was genetically stable following infection in mice , though secondary site mutations arise adjacent to the CD-loop . Importantly , while shortening of the CD-loop attenuates the virus , the CD-loop mutant maintains antigenicity in immunocompetent mice , eliciting an antibody response that similarly neutralizes both the mutant and wildtype ZIKV . These findings suggest that the extended CD-loop in ZIKV is a determinant of neurotropism and may be a target in live-attenuated vaccine design , for not only ZIKV , but for other neurotropic flaviviruses .
Zika virus ( ZIKV ) is a positive-sense single-stranded RNA flavivirus . Though first isolated in 1947 and after years of relatively benign infections throughout Southeast Asia , ZIKV was identified in large human disease outbreaks in Yap Island , French Polynesia , and then the Americas [1–5] . ZIKV infection is asymptomatic in a majority of adult cases , but when symptomatic , generally causes mild febrile illness [6 , 7] . ZIKV infection has also been associated with more severe disease , such as neurological complications including Guillain-Barré syndrome [7–10] . Additionally , ZIKV infection during pregnancy is linked to microcephaly and fetal demise [11 , 12] . The ability of ZIKV to cause neurological disease is not unique among flaviviruses [13] . West Nile virus ( WNV ) , Japanese encephalitis virus ( JEV ) , and tick-borne encephalitis virus ( TBEV ) also infect the central nervous system ( CNS ) and cause encephalitic disease . However , dengue virus ( DENV ) , a highly similar virus both genetically and antigenically , only very rarely causes neurological or encephalitic disease [14 , 15] , highlighting a disconnect between disease and virus relatedness . While the complete viral determinants of flavivirus neurotropism are not fully known , there are several described conserved and divergent mechanisms [13] . For example , multiple groups have shown that envelope protein glycosylation is necessary for neurovirulence of WNV and ZIKV [16–19] . However , this glycosylation motif is conserved between both neurotropic and non-neurotropic flaviviruses , indicating that it is not solely sufficient for neurotropism . Additionally , several groups have identified novel neuronal ZIKV receptors , such as AXL , that are not utilized by other neurotropic flaviviruses [20 , 21] . Multiple groups have shown that ZIKV is more structurally stable than DENV , hypothesizing that it can persist longer in body compartments and fluids , potentially leading to an increased chance of neuroinvasion [22–25] . One key difference between the structural envelope proteins of neurotropic and hemorrhagic flaviviruses is the extension of the CD-loop by a single amino acid [23 , 24] . Though predicted to stabilize the virus via a network of hydrogen bonds , we have previously shown that the extended loop itself , independent of hydrogen bonding , is responsible for ZIKV’s structural and thermal stability [23 , 24] . Additionally , shortening of the CD-loop by a single amino acid ( Δ346 ZIKV ) attenuated the virus in a mouse model of ZIKV pathogenesis [24] . Interestingly , while both viruses replicate in the periphery , the CD-loop mutant was less likely to be found in the CNS of infected mice compared to wildtype ( WT ) ZIKV . This suggests that the extended CD-loop , or the structural stability that it confers , is important for neuropathogenesis . In this study we further investigate the role of the ZIKV CD-loop in neurotropism and antigenicity . We find that the CD-loop mutant is delayed in disseminating to the brain , but once present can replicate and cause lethal disease . Importantly , when delivered via intracranial infection , Δ346 ZIKV is still less pathogenic , indicating that the attenuated phenotype is not solely due to a defect in neuroinvasion . While the Δ346 deletion is genetically stable after long-term in vivo infection , secondary site mutations emerge in the envelope glycoprotein . A successful ZIKV vaccine should have no risk of neurological complications in immunocompetent populations , and as our shortened CD-loop virus results in decreased neuropathogenicity , we characterized the antibody response elicited by the Δ346 virus in mice . Mice infected with the Δ346 mutant mount an antibody response that similarly neutralizes both Δ346 and WT ZIKV in cell culture . Together , this data suggests that shortening the CD-loop of ZIKV results in decreased neurovirulence while maintaining wildtype ZIKV antigenicity .
We have previously shown that shortening of the ZIKV CD-loop results in attenuated replication in both mammalian Vero and mosquito C6/36 cells [24] . To determine if the Δ346 virus is attenuated in neuronal cells , we performed a single step growth curve in the human neuroblastoma cell line SH-SY5Y , which have been previously used as an in vitro neurological model of ZIKV replication ( Fig 1 ) [26–28] . Though Δ346 ZIKV replicated similarly to WT ZIKV at early time points , the Δ346 virus was slightly attenuated throughout most of the time course before finally reaching equivalent peak titers ( p<0 . 05 ) . This indicates that despite minor attenuation relative to WT ZIKV , the CD-loop mutant is capable of infection and replication in neuronal cells . We previously explored the Δ346 mutant in 11-week-old IFNAR-/- IFNGR-/- mice on a C57BL/6J background ( herein referred to as IFNAGR-/- ) , observing severe attenuation with no development of disease [24] . Younger mice are more susceptible to ZIKV disease [29] , allowing us to study the Δ346 mutant in a more permissive model . To determine if the Δ346 ZIKV pathogenesis attenuation also occurs in younger mice , we infected 8- to 10-week-old IFNAGR-/- mice and monitored weight loss , lethality and viral loads in various tissues ( Figs 2–3 ) . Mice infected with WT ZIKV all succumb to disease by day 15 post infection , whereas lethality is delayed in Δ346 ZIKV infected mice ( p<0 . 05 ) ( Fig 2A and 2B ) . Increasing the infection dose of Δ346 ZIKV from 103 to 104 FFU did not increase lethality or weight loss , indicating that the attenuation phenotype is not directly dose-dependent ( p>0 . 05 ) ( Fig 2C and 2D ) . On day six post infection , the Δ346 mutant replicated to lower titers in the serum , liver , spleen , and kidneys compared to WT virus ( p<0 . 05 ) ( Fig 3A ) . Additionally , at this early time point , there were fewer Δ346-infected mice with detectable virus in their brain and spinal cord compared to WT ZIKV infected mice ( p < 0 . 05 ) ( Fig 3B ) . IFNAGR-/- mice typically succumb to ZIKV infection due to neurological complications caused by viral replication in the brain [29] . Moribund infected mice were harvested for organ end point virus titers . Despite succumbing to disease at different rates ( Fig 2A ) , nearly all WT and Δ346 ZIKV infected mice cleared serum viremia by time of death ( Fig 3C ) . Interestingly , mice infected with either virus had comparable titers in the brain and spinal cord ( p>0 . 05 ) ( Fig 3C ) , demonstrating that although the Δ346 virus is attenuated , it can eventually replicate to high titers and cause encephalitic disease . Overall , shortening of the CD-loop results in decreased dissemination of ZIKV to the brain and delayed mortality , but it still causes neurological disease in immunocompromised mice . To determine if the Δ346 virus is genetically stable in IFNAGR-/- mice , we sequenced the envelope gene of virus from the brains of mice infected via footpad with WT and Δ346 ZIKV . All mice infected with the Δ346 mutant with detectible virus in their brain ( Fig 3B and 3C ) maintained the deletion of residue 346 , demonstrating that this deletion is stable in vivo ( Fig 4A ) . However , viruses in these mice had envelope protein secondary site mutations ( Fig 4A ) that were not detected in the virus stock ( sequence available in S1 Data ) . At day six post infection , 100% of Δ346 ZIKV virus-positive mice contained a V391A substitution . By the end point , virus from 60% of Δ346 ZIKV infected mice contained this same substitution , indicating this residue might be important in dissemination to or replication in the CNS . Viruses from individual mice also contained three other substitutions , however each of these mutations were only present in virus from a single mouse , suggesting they are not common mutations . These secondary site mutations , with the exception of G150E , are all structurally located near the CD-loop , suggesting that they may have emerged as a consequence of the shortened CD-loop ( Fig 4B and 4C ) . Importantly , none of these mutations emerged in mice infected with WT ZIKV ( Fig 4A ) , revealing that these substitutions arise specifically in the context of Δ346 ZIKV . It is possible that these secondary site mutations are functionally compensating for decreased stability of the Δ346 or decreased fitness in the CNS . Future studies are necessary to elucidate the precise impact of these mutations . To determine if Δ346 ZIKV is able to replicate similarly to WT ZIKV in the brain , without the barrier of dissemination into the CNS , we performed intracranial ( IC ) infections in IFNAGR-/- and wildtype C57BL/6J mice . Following IC injection , immunocompromised mice infected with the Δ346 mutant had delayed lethality relative to WT virus ( p<0 . 05 ) ( Fig 5A and 5B ) , but succumb to infection more rapidly than when inoculated via footpad infection ( Fig 2A and 2B ) . CD-loop mutant infected mice have similar brain viral titers on days two and six compared , and end point to WT ZIKV infected mice ( Fig 5C ) . Wildtype C57BL/6J mice did not show any signs of disease or viral replication in the CNS following IC infection with either virus ( Fig 5D–5F ) . Even when delivered to the brain , an immunopriveledged site , of immunocompetent mice , neither virus is capable of causing disease . Overall , the shortened CD-loop mutant is attenuated relative to WT ZIKV at replicating the CNS , causing delayed disease in immunocompromised mice . As the Δ346 mutant is attenuated in neuropathogenesis , mutations that shorten the ZIKV CD-loop may be incorporated into future live-attenuated vaccines . We have previously shown that while the Δ346 virus is structurally destabilized , viruses produced in cell culture are morphologically similar to WT ZIKV [24] . To determine if Δ346 ZIKV infection elicits an antibody response similar to that of WT ZIKV infection , we infected C57BL/6J mice pretreated with an IFNAR-blocking antibody and analyzed the anti-ZIKV antibodies . Mice infected with WT or Δ346 ZIKV did not show any signs of disease , but supported acute viral replication with similar serum viremia on day three post infection ( p>0 . 05 ) ( Fig 6A and 6B ) . We next evaluated the ability of WT and Δ346 ZIKV immune sera to bind and neutralize both WT and Δ346 ZIKV viruses . WT ZIKV immune sera had similar levels of ZIKV-reactive IgG compared to Δ346 sera , and both sera bound both viruses similarly ( p>0 . 05 ) ( Fig 6C ) . Despite lower levels of total ZIKV antibodies , sera from Δ346-infected mice efficiently neutralized both viruses , with neutralization titers similar to those of WT ZIKV sera ( p>0 . 05 ) ( Fig 6D ) . Importantly , these data show that the Δ346 mutant maintains antigenicity and elicits neutralizing antibodies against WT ZIKV suggesting potential application as a part of a live-attenuated vaccine approach .
Having previously characterized the extended ZIKV CD-loop as an important determinant of stability and virulence [24] , here we further examined the shortened CD-loop mutant virus . The Δ346 mutant displays a mild growth defect in a neuronal cell line suggesting capacity to replicate in the brain . In addition , we confirmed in vivo attenuation via the peripheral route of infection in terms of dissemination to the CNS as well as replication in the brain relative to WT ZIKV . The Δ346 mutation is genetically stable after in vivo infection . However , Δ346 virus sequenced from the brains of mice reveal secondary site mutations near the CD-loop whose functions are unknown , but may compensate for viral stability or neuropathogenicity . The secondary site mutations identified in this study may inform future structure-function studies of the CD-loop in regards to ZIKV virion stability and pathogenesis , identifying key CD-loop contacts critical in the ZIKV virion . After intracranial infection , the Δ346 mutant was highly attenuated , demonstrating that in addition to defects in neuroinvasion , the CD-loop mutant causes delayed encephalitic disease when delivered directly to the CNS . While the CD-loop mutant displays attenuation , it is capable of eliciting antibodies that neutralize WT ZIKV , suggesting it still maintains proper ZIKV antigenicity . While we do not know the mechanism of the Δ346 mutant attenuation beyond virion stability , the shortened CD-loop may alter cellular tropism , particularly in the brain , resulting in fewer infected cells and delayed disease . Defining the cell types infected by the Δ346 mutant and WT ZIKV in tissues , specifically the brain , may provide insights into the decreased disease phenotype . The viral determinants that allow ZIKV and other neurotropic flaviviruses to invade and replicate in the CNS are not fully understood [6 , 7 , 13] . Here , we show that shortening of the CD-loop in the ZIKV envelope results in delayed neuroinvasion in immunocompromised mice . This delayed neuroinvasion may be due stochastic differences in intra-host evolution of the Δ346 virus developing secondary site mutations that increase fitness in the CNS . Neurotropic flaviviruses WNV , JEV , and TBEV also share an extended CD-loop within the envelope protein relative to hemorrhagic flaviviruses such as DENV [23] . Shortening the CD-loop in other flaviviruses would be important in understanding if the CD-loop modulates stability and neurovirulence in other viruses . Within ZIKV strains , there is variation in the pathogenicity of ZIKV isolates . Multiple groups have shown that Asian ZIKV strains , such as the H/PF/2013 strain used in this study , are less neurovirulent than African strains [30 , 31] . The extended CD-loop is conserved among all ZIKV strains , thus we expect that shortening the CD-loop in African ZIKV strains will also attenuate the virus , though the effects may be more modest due to other determinants driving their increased neuropathogenicity . Targeting the CD-loop may be a broadly applicable approach for flavivirus attenuation . Groups have shown that glycosylation of the WNV envelope protein has been well studied for its role in neurotropism , however , it is possible the extended CD-loop is an additional viral determinant of neurotropism [16 , 17] . Although speculative , a WNV CD-loop mutant could be an important tool for studying CD-loop-mediated neurotropism , as WNV causes encephalitic disease in immunocompetent mice [16 , 17] . Relative to DENV , both ZIKV and WNV have an extended CD-loop; however , the identity of residue 346 varies , with alanine in ZIKV and valine in WNV . It remains unclear whether the length of the CD-loop or the identity of the residue at position 346 is critical for pathogenesis . While several flavivirus vaccines exist , the YFV vaccine is occasionally associated with rare complications [32–35] . Though YFV is non-neurotropic , it contains an extended CD-loop [23] and incorporation of the Δ346 mutation into the YFV vaccine may further increase safety , while preserving antigenicity and efficacy . Live-attenuated vaccines resulting from deletions , such as a shortened CD-loop , are also more likely to be genetically stable than attenuating substitution mutations . There are limitations of studying ZIKV pathogenesis in current animal models , including the necessity of using immunocompromised mice due to the strong restriction of ZIKV by murine interferon signaling [29 , 36] . Circumventing the interferon restriction , via genetic knockout mice or antibody blockade of the interferon response , permits ZIKV replication , but limits our ability to study the intact immune response to infection . Additionally , the use of immunocompromised mice with delayed viral clearance is anticipated to increase the likelihood of accumulating secondary site mutations which may contribute to restored replication fitness and pathogenesis . Mouse adapted ZIKV viruses and use of humanized or human STAT2 knock-in mice may serve as better future models to understand innate and adaptive responses [37] . As these models become more available , it will be important to study the kinetics , pathogenesis , and genetic stability of Δ346 ZIKV in these more relevant , immunocompetent models . These models will also be important in testing the genetic stability of Δ346 ZIKV in an immunocompetent model . Overall , this work highlights the importance of the extended CD-loop for ZIKV neuropathobiology . The extended CD-loop is an important determinant for ZIKV dissemination and CNS invasion , but also crucial for efficient replication in the brain . However , shortening of the CD-loop does not abolish neuropathogenicity , highlighting the multifaceted nature of ZIKV neurovirulence . Additionally , shortening of the ZIKV CD-loop , while attenuating the virus , maintains antigenicity and may be utilized in next-generation vaccine design in combination with additional attenuating mutations in the genome that reduce chance of disease .
Viruses were generated using a ZIKV H/PF/2013 infectious clone as described previously [24 , 38] . Low passage virus stocks were produced in C6/36 cells . C6/36 cells ( ATCC CRL-1660 ) were maintained in Minimum Essential Medium , supplemented with 5% fetal bovine serum ( FBS ) , 1% nonessential amino acids ( NEAA ) , and 1% antibiotic-antimycotic ( AA ) and grown in 5% CO2 at 32°C . SH-SY-5Y cells ( ATCC CRL-2266 ) were maintained in MEM:F12 supplemented with 10% FBS and 1% AA and grown in 5% CO2 at 37°C . Samples were titered on C6/36 cells as the particle to FFU ratio is more similar between WT and Δ346 ZIKV than when titered on Vero cells [24] . Viruses were serially diluted and added to C6/36 monolayers for one hour . Cells were then overlaid with OptiMEM supplemented with 1% methylcellulose , 2% FBS , 1% NEAA , and 1% AA and incubated at 32°C for 4–5 days . Cells were gently washed with PBS and fixed in cold 50% acetone + 50% methanol , then immunostained with either mouse anti-E MAb 4G2 or human anti-E MAb 1M7 , HRP-labeled secondary antibody , and developed with KPL TrueBlue substrate . Plates were seeded with SH-SY5Y cells one day prior to infection . Cells were infected in triplicate at an MOI of 1 for two hours , then washed with PBS to remove unbound virus , and replaced with fresh media . Every 24 hours , supernatant was sampled , and immediately frozen at -80°C . Samples were titered on C6/36 cells as described above . Two-factor ANOVA followed by Sidak’s multiple comparison was conducted on log-transformed data . Experiment was performed once . This study was carried out in accordance with the recommendations for care and use of animals by the Office of Laboratory Animal Welfare ( OLAW ) , National Institutes of Health . All animal work was performed in strict adherence to the Institutional Animal Care and Use Committee ( IACUC ) and University of North Carolina at Chapel Hill policy . The protocol [#16–090] was approved by the UNC’s IACUC ( Permit Number A-3410-01 ) . Footpad infections were performed in 8- to 10-week-old male and female IFNAGR-/- mice on C57BL/6 background with 103 or 104 FFU of WT or Δ346 ZIKV in 10ul diluted in PBS into the left hind footpad ( n = 13 for each group ) . Infected mice were monitored daily for weight loss and clinical signs of disease . On day six post infection , a subset of mice were euthanized and perfused with PBS for tissue analysis . The remainder of mice were euthanized via isoflurane overdose and cervical dislocation upon losing 20% of their starting body weight or showing signs of severe disease and tissues were collected for titer analysis . No mice died prior to humane euthanasia . Tissues were homogenized in 1ml PBS , centrifuged to pellet debris , immediately frozen at -80°C , then clarified supernatant was used for virus titering as described above . Survival curves were analyzed by log-rank test . Log-transformed titer data was analyzed by Mann-Whitney test . Intracranial infections were performed in 5- to 7-week-old male and female C57BL/6J or IFNAGR-/- mice with 103 FFU of WT or Δ346 ZIKV in 20ul diluted in PBS into the left cerebral hemisphere via insulin syringe while under brief isoflurane anesthesia and monitored as described above ( For IFNAGR-/- mice , n = 11 and 14 respectively . For C57BL/6J mice , n = 15 and 14 , respectively ) . Mice were euthanized on days two , six , or upon losing 20% of their starting weight , and tissue was collected and titered as described above . No mice died prior to humane euthanasia . Survival curves were analyzed by log-rank test . Log-transformed titer data was analyzed by two-factor ANOVA followed by Sidak’s multiple comparisons . Antibody response experiments were performed in five-week-old female C57BL/6J mice , which were given 1mg of anti-IFNAR1 antibody ( MAR1-5A3 , Bio X Cell ) via intraperitoneal injection one day prior to footpad infection and monitored as described above ( n = 5 mock infected and n = 8 for each ZIKV infected group ) . Blood was collected on day three post infection via submandibular bleed , allowed to clot for at least 10 minutes and clarified by centrifugation prior to storage at -80°C . Mice were euthanized on day 28 post infection for terminal blood collection . Log-transformed titer data was analyzed by one-factor ANOVA followed by Tukey’s multiple comparisons . All mouse experiments were performed once . Virus stock RNA was isolated via Viral RNA Mini Kit ( Qiagen ) . Total brain RNA was isolated from brain homogenates and TRIzol LS ( Invitrogen ) and Direct-zol RNA MiniPrep kit ( Zymo Research ) . cDNA was reverse transcribed using SuperScript III ( Invitrogen ) using random primers . The ZIKV envelope sequence was PCR amplified , and PCR amplicon was sequenced via Sanger sequencing and analyzed in Geneious ( Version 11 . 0 . 04 ) . 96-well high bind plates were coated with human MAb EDE1 C10 , previously shown to bind WT and Δ346 ZIKV equally [24] . Plates were blocked with 3% non-fat milk , then viral antigen ( WT or Δ346 ZIKV ) was captured for one hour at 37°C . Mouse immune sera was diluted 1:1000 then allowed to bind to captured virus for one hour at 37°C . Anti-mouse-IgG alkaline phosphatase labeled secondary antibody was added , plates were developed using p-nitrophenyl phosphate , and color changes were quantified via spectrophotometry . ELISA data was analyzed by two-factor ANOVA followed by Tukey’s multiple comparisons . Plates were seeded with C6/36 cells one day prior to neutralization assay . Mouse immune sera were serially diluted four-fold beginning at 1:20 , then mixed with WT or Δ346 ZIKV viruses diluted to ~45 FFU/well . Virus:Ab mixture was incubated for one hour at 32°C , added to C6/36 cells and incubated for an additional hour at 32°C . Overlay was added and cells were incubated for 4–6 days , then fixed and immunostained as described above . FRNT data was analyzed by two-factor ANOVA followed by Tukey’s multiple comparisons .
|
Zika virus ( ZIKV ) is a mosquito-transmitted virus that was recently introduced in Brazil and subsequently spread throughout the Americas . ZIKV is highly similar to the related dengue virus but causes unique disease outcomes including neurological disease in adults and fetal developmental complications . The ZIKV envelope protein coats the surface of the virus and allows entry into host cells . Here we investigate a portion of the ZIKV envelope protein , the CD-loop , which extends further than in dengue virus , and its role in ZIKV neurological disease . Our study finds that shortening the CD-loop reduces the ability of ZIKV to replicate in neuronal cells , that the longer CD-loop is a key factor for invasion of the central nervous system in mice , and that a deletion in the CD-loop is genetically stable with passage . Additionally , we show that infection with the CD-loop mutant induces a potent antibody response that can neutralize wildtype ZIKV , suggesting it may offer protection in mice . Shortening of the ZIKV CD-loop , and the CD-loop of other neurotropic flaviviruses , could contribute to development of rapid , effective , and safe vaccines .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"microbial",
"mutation",
"enzyme-linked",
"immunoassays",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"nervous",
"system",
"pathogens",
"immunology",
"microbiology",
"viruses",
"mutation",
"substitution",
"mutation",
"rna",
"viruses",
"antibodies",
"immunologic",
"techniques",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"immunoassays",
"viral",
"replication",
"biochemistry",
"anatomy",
"flaviviruses",
"central",
"nervous",
"system",
"virology",
"viral",
"pathogens",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"organisms",
"zika",
"virus"
] |
2019
|
Shortening of Zika virus CD-loop reduces neurovirulence while preserving antigenicity
|
Infection with Plasmodium vivax results in strong activation of monocytes , which are important components of both the systemic inflammatory response and parasite control . The overall goal of this study was to define the role of monocytes during P . vivax malaria . Here , we demonstrate that P . vivax–infected patients display significant increase in circulating monocytes , which were defined as CD14+CD16− ( classical ) , CD14+CD16+ ( inflammatory ) , and CD14loCD16+ ( patrolling ) cells . While the classical and inflammatory monocytes were found to be the primary source of pro-inflammatory cytokines , the CD16+ cells , in particular the CD14+CD16+ monocytes , expressed the highest levels of activation markers , which included chemokine receptors and adhesion molecules . Morphologically , CD14+ were distinguished from CD14lo monocytes by displaying larger and more active mitochondria . CD14+CD16+ monocytes were more efficient in phagocytizing P . vivax-infected reticulocytes , which induced them to produce high levels of intracellular TNF-α and reactive oxygen species . Importantly , antibodies specific for ICAM-1 , PECAM-1 or LFA-1 efficiently blocked the phagocytosis of infected reticulocytes by monocytes . Hence , our results provide key information on the mechanism by which CD14+CD16+ cells control parasite burden , supporting the hypothesis that they play a role in resistance to P . vivax infection .
Plasmodium vivax is the most widely distributed malaria parasite and responsible for approximately 70–80 million cases , annually . In addition , P . vivax is responsible for the majority of malaria cases and represents a significant impediment to social and economic development in Latin America and Asia [1] . Both innate and acquired immunity are thought to play critical role in host resistance to infection and pathogenesis of malaria [2] , [3] . However , the mechanisms by which the innate immune response mediate resistance to Plasmodium infection or promote a deleterious systemic inflammation associated with malaria sepsis are poorly understood [2] . This is particularly true in the case of P . vivax malaria [4] . The blood is the main tissue affected during P . vivax malaria since sequestration is not a central event in this infection . When parasitized reticulocytes rupture in the blood stream , parasite components are sensed by the innate immune receptors and activate monocytes [5] . The innate immune system recognizes Plasmodium sp . through different pattern-recognition receptors expressed by monocytes and initiates a broad spectrum of defense mechanisms [6] , [7] , [8] , [9] , [10] . Importantly , the same immune mediators involved in host resistance , such as pro-inflammatory cytokines are also thought to mediate pathology during acute malaria episodes [8] , [11] , [12] . However , the full spectrum of monocyte subsets and the specific functions of each monocyte population during malaria have not been defined . Besides supplying peripheral tissues with macrophage and dendritic cell ( DC ) precursors , monocytes contribute directly to immune defense against microbial pathogens [13] , [14] , [15] . Monocytes were initially identified by their expression of large amounts of CD14 [16] , [17] . However , recent studies have revealed that monocytes in human peripheral blood are heterogeneous and can be divided into three distinct subsets described based on their expression of phenotypic markers . These cells are referred to as , classical ( CD14+CD16− cells ) , inflammatory or intermediate ( CD14+CD16+ ) and patrolling ( CD14loCD16+ ) monocytes [18] , [19] . Given the importance of monocytes as a major source of pro-inflammatory cytokines and potential effector cells during malaria , in this study , we have attempted to define the role of the various monocyte subsets during P . vivax infection . To address this question , we phenotypically and functionally characterized the classical , inflammatory , and patrolling monocytes present in the peripheral blood from patients experiencing acute malaria episodes . We demonstrate that the frequency of circulating monocytes is elevated during acute infection with P . vivax and that the classical and inflammatory monocyte subsets are the primary source of pro-inflammatory cytokines . Importantly , we found that CD16+ cells , and in particular the CD14+CD16+ LFA-1highICAM-1highPECAM-1high monocytes display augmented effector functions such as phagocytosis and intracellular reactive oxygen species production and are thus likely to be an important cell subset controlling parasitemia , and mediating host resistance to P . vivax infection .
A total of 35 P . vivax-infected patients with uncomplicated malaria were enrolled in this study at Centro de Pesquisa de Medicina Tropical de Rondônia ( CEPEM ) in Porto Velho , Rondônia , a malaria endemic area in the Amazon region of Brazil . According to the World Health Organization , uncomplicated malaria is a symptomatic infection with malaria parasitemia without signs of severity and/or evidence of vital organ dysfunction . Up to 100 ml of peripheral blood was collected immediately after confirmation of P . vivax infection by thick blood smear film and again 30–45 days after chemotherapy ( n = 35 , ranging from 18 to 66 years old [35±9 . 5] ) ( Table S1 ) . Additional 36 P . vivax-infected patients were enrolled for reticulocyte analysis: percentage: mean: 1 . 383 , SD: 0 . 655; median: 1 . 150 , IQR: 0 . 925–1 . 650; absolute numbers: mean: 51 , 779 , SD: 18 , 390; median: 48 , 100 , IQR: 38 , 970–59 , 520 . Patients were treated for 3 days with chloroquine followed by 7 days with primaquine according to the Brazilian Ministry of Health . P . vivax infection and parasitological cure were confirmed by thick blood smear film and polymerase chain reaction ( PCR ) [20] . Identification of the three species of human malaria parasites was done by nested PCR that targets variant sequences in the small subunit rRNA gene . The clinical manifestations of acute malaria were fever , myalgia , chills , arthralgia , nausea , vomiting or diarrhea . Peripheral blood was also collected from 15 healthy donors ( HD ) ranging from 21 to 56 years old [32±8] living in Porto Velho and negative for P . vivax infection . These studies were performed under protocols reviewed and approved by the Ethical Committees on Human Experimentation from Centro de Pesquisa em Medicina Tropical de Rondônia ( CEP-CEPEM 095/2009 ) and Centro de Pesquisas René Rachou , Fundação Oswaldo Cruz ( CEP-CPqRR 2004 ) , the National Ethical Committee ( CONEP 15652 ) from Ministry of Health , Brazil , as well as by the Institutional Review Board from the University of Massachusetts Medical school . Only adults , 18 years old , were enrolled in the study . All patients enrolled in this study provided written informed consent . Peripheral blood mononuclear cells ( PBMC ) were prepared from heparinized venous blood of adult volunteers by Ficoll-Hypaque density gradient centrifugation ( GE Healthcare Life Sciences ) . Cells were stained for surface molecules for 15 minutes at room temperature . Subsequently , PBMC were washed with FACS buffer ( PBS and 2%FBS ) , and fixed and permeabilized according to the manufacture's instruction ( Cytofix/Cytoperm , BD Biosciences ) . Cells were then stained for the intracellular antigens . Cells were finally suspended and maintained in 200 µl of PBS 2% paraformaldehyde ( Sigma Aldrich ) . At least 100 , 000-gated events were acquired for analysis using FACSCan upgraded with a second laser ( 5 colors ) with Cellquest Pro and Rainbow from Cytek or LSR II with Diva ( both from BD Biosciences ) . The antibody panels included the following: anti-CD31 ( clone WM59 ) -FITC , anti-CD11a ( LFA-1 ) ( clone HI111 ) -FITC , anti-HLA-DR ( clone LN3 ) -PE , anti-CD14 ( clone 61D3 ) -APC , anti-CD62L ( clone DREG56 ) -APC , anti-HLA-DR ( clone LN3 ) -efluor 780 , anti-CX3CR1 ( clone 2A9-1 ) -biotin , anti-CD106 ( VCAM-1 ) ( clone STA ) -biotin and streptavidin-PE , all purchased from eBioscience; anti-CD54 ( ICAM-1 ) ( clone HA58 ) -PE , anti-TNF-α ( clone MAb11 ) -PE , anti-IL-6 ( clone MQ2-13A5 ) -PE , anti-CD197 ( CCR7 ) ( clone 3D12 ) -PECy7 and anti-CD16 ( clone 3G8 ) -PercPCy5 . 5 , purchased from BD Biosciences; anti-CCR2 ( clone 48607 ) -PE , purchased from R&D Systems . Data were analyzed using FlowJo Version 9 . 3 . 2 ( TreeStar ) . A forward scatter height ( FSC-H ) and a side scatter height ( SSC-H ) gate were used to initially remove debris and to capture leucocytes . CD14 versus SSC-H gate was then used to select monocytes . An additional HLA-DR versus CD16 gate was performed to exclude possible contamination by neutrophils ( CD16+HLA-DR− ) . A more detailed analysis of monocyte subpopulations was done based on CD14 and CD16 expression and here designated as: classical monocytes ( CD14+CD16− ) , inflammatory monocytes ( CD14+CD16+ ) , and patrolling monocytes ( CD14l°CD16+ ) . The activation/cell presentation molecule , HLA-DR; the cellular adhesion molecules , VCAM-1 , ICAM-1 , PECAM-1 , and LFA-1; the chemokine receptors , CCR2 , CX3CR1 and CCR7; and the cytokine TNF-α was analyzed within CD14+ cells and also within each monocyte subpopulation . Data is shown in frequencies or mean fluorescence intensity ( MFI ) . In the latter , when graphs are overlaid , the y-axis is left on automatic scaling and the axis represents % of Maximum . This normalization is used because different numbers of events is collected for the monocyte subsets analyzed and allow us to focus on the important and relevant variations between the levels of expression of different markers on the x-axis . IL-6 , IL-8 and IL-10 were measured in cryopreserved plasma using the Cytometric Bead Array kit ( CBA , BD Biosciences Pharmingen ) as recommended by the manufacturer . The concentration of cytokines in each sample was calculated using the BD FCAP Array Software v 1 . 0 . 1 ( BD Biosciences ) . The red blood cells pellet from the Ficoll-Hypaque ( GE Healthcare Life Sciences ) density gradient centrifugation was harvested and washed three times and then resuspended in RPMI 1640 medium ( Sigma Aldrich ) to a final hematocrit of 10% . Five milliliters of this suspension was overlaid on a 5 mL 45% Percoll solution in a 15 mL tube . After centrifugation , floating mature Pv-reticulocytes ( Pv-Ret ) were collected and resuspended in RPMI 1640 [21] , [22] . The enrichment of Pv-Ret was assessed by optic microscopy and a purity of 95% was obtained . Where indicated , enriched Pv-Ret were stained with 1 µM carboxyfluorescein succinimidyl ester ( CFSE ) at 1×106 cells/mL for 8 minutes at room temperature before phagocytosis assays ( Molecular Probes-Invitrogen ) . After stained Pv-Ret were washed three times in RPMI and 10%FBS . P47phox and p67phox expression was assessed by incubating PBMC in suspension at 37°C in complete RPMI 1640 supplemented with 2 µM glutamine , 10 mM HEPES and 50 µM 2-ME for 3 hours in medium alone and with P . vivax-Ret ( 0 . 5 Pv-Ret/PBMC ) in the presence of 10% immune serum . PBMC were then washed with FACS buffer , stained for surface molecules for 15 minutes at room temperature , fixed and permeabilized according to the manufacture's instruction ( Cytofix/Cytoperm , BD Biosciences ) . . Cells were then stained for the intracellular antigens ( p47phox and p67phox ) for 20 minutes , following additional 20 minutes incubation with anti-IgG1 . After staining , plates were kept on ice for 15 minutes and cells were harvested with ice-cold PBS , 2 . 5 mM EDTA and maintained in 200 µl of PBS 2% paraformaldehyde . At least 100 , 000-gated events were acquired for analysis using a LSRFortessa ( BD Biosciences ) . The antibody panels included the following: anti-CD16 ( clone 3G8 ) -Alexa Fluor 700 , anti-HLA-DR ( Tu39 ) -FITC , anti-CD14 ( clone M5E2 ) -APC , purified anti-phox47 ( 1/p47Phox ) and anti-phox67 ( D-6 ) , and anti-IgG1 ( A851 ) -PE . The MFI of each monocyte subpopulation expressing the NADPH subunits was determined by flow cytometry . Data were analyzed using FlowJo Version X 10 . 0 . 7 . Phagocytosis was assessed by incubating PBMC in suspension using 96-wells polystyrene plates at 37°C in complete RPMI 1640 supplemented with 10% heat-inactivated FCS , 2 µM glutamine , 10 mM HEPES , and 50 µM 2-ME for 1 , 4 and 12 hours with P . vivax-Ret previously stained with CFSE ( 0 . 5 Pv-Ret/PBMC ) in the absence of immune serum and in the presence of serum or inactivated serum . In some experiments monoclonal anti-CD11a ( 1 mg/mL ) ( clone G43-25B , BD ) , anti-CD31 ( 0 . 5 mg/mL ) ( clone WM59 ) , anti-CD36 ( 0 . 5 mg/mL ) ( clone CB38 ) and anti-CD54 ( 1 mg/mL ) ( clone HA58 ) blocking antibodies were added in the cultures 30 minutes before the addition of P . vivax-Ret . After staining , plates were kept on ice for 15 minutes and cells were harvested with ice-cold PBS containing 2 . 5 mM EDTA . The frequencies of total monocytes and each monocyte subpopulation positive for CFSE were determined by flow cytometry and data were analyzed using FlowJo Version 9 . 3 . 2 ( TreeStar ) . To detect ROS production at the single cell level , the Image-iT LIVE Green Reactive Oxygen Species Detection kit ( Invitrogen ) was used following the manufacturer's instructions . Briefly , PBMC were washed twice with PBS and incubated in medium alone or with Pv-Ret ( 0 . 5 Pv-Ret/PBMC ) or phorbol 12-myristate 13-acetate ( PMA , 10 ng/mL ) and ionomycin ( 500 ng/mL ) . Pre-warmed 25 µM 5- ( and-6 ) -carboxy-2′ , 7′-dichlorodihydrofluorescein diacetate ( carboxy-H2DCFDA ) was added to the cells for 3 hours at 37°C . Mitochondrial ROS was assessed by MitoSox red mitochondrial superoxide indicator ( Invitrogen ) following the manufacturer's instructions . Briefly , PBMC were washed twice with HBSS and incubated in medium alone or with Pv-Ret ( 0 . 5 Pv-Ret/PBMC ) or PMA ( 10 ng/mL ) and ionomycin ( 500 ng/mL ) and 10 µm MitoSox . After 30 minutes incubation , monoclonal antibodies against CD14 and CD16 were added to cell cultures to allow monocyte subpopulations analysis . Cell suspension was washed after additional 30 minutes incubation with HBSS at 37°C . Plates were kept on ice for 15 minutes and cells were harvested with ice-cold PBS containing 2 . 5 mM EDTA . Cells were acquired by flow cytometry and data were analyzed using FlowJo Version 9 . 3 . 2 . After PBMC preparation CD14+CD16− , CD14+CD16+ and CD14loCD16+ monocytes from HD and P . vivax-infected patients were sorted with a FACSAria II cell sorter ( BD Biosciences ) , using the combination of antibodies described above . CD14+CD16− , CD14+CD16+ and CD14l°CD16+ monocytes were then collected and fixed with 2 . 5% buffered glutaraldehyde solution , 0 . 1 M , for electron microscopy or with RLT buffer ( QIAGEN ) supplemented with β-mercaptoethanol for mRNA detection and nanostring analysis as described below . After FACS-sorting , cells were prepared as previously described [23] , [24] . Briefly , cells were fixed in 2 . 5% buffered glutaraldehyde solution , 0 . 1 M , pH 7 . 2 , 6 h , 8°C . Cells were then washed with the same buffer . The pellets were included in phosphate buffer , 4% agarose and left overnight at 4°C . Next , the cells were fixed in a mixture of 1% osmium tetroxide and 1 . 5% ( w/v ) potassium ferrocyanide , dehydrated in a graded series of ethanol solutions , infiltrated , and embedded in Araldite 502 ( Electron Microscopy Sciences , Hatfield , PA , USA ) . After polymerization , thin sections were obtained using a diamond knife on a Sorvall MT-2B ultramicrotome ( Dupont , Wilmington , DE , USA ) and mounted on uncoated 200-mesh copper grids ( Ted Pella , Inc . , Redding , CA , USA ) . Sections were stained with 2% uranyl acetate and Reynolds lead citrate and then analyzed using transmission electron microscopy ( EM 10A Zeiss ) . mRNA was assessed by nanostring analysis [25] . nCounter CodeSets were constructed for detecting selected human-specific genes . A total of 1×104 cells of each subset were lysed in RLT buffer ( QIAGEN ) supplemented with β-mercaptoethanol . This lysate was mixed with capture and reporter probes , hybridized to the Codeset for 16 hr and loaded onto the nCounter prep station , and then quantified with the nCounter Digital Analyzer . For side-by-side comparisons of nCounter experiments , data were normalized in two ways described previously [26] . Briefly , the first normalization was for small variations utilizing the internal positive controls that are present in each CodeSet . Then the samples were normalized with 7 housekeeping genes that were included in the CodeSet . The data was analyzed with n Solver software . The heatmap was constructed using log2 transformed data and the Tiger Multi Experiment Viewer software . Statistical analysis was performed using GraphPad Prism software , version 5 . 0 . The results were analyzed using two-tailed paired t-test . Wilcoxon testing was used when data did not fit a Gaussian distribution . The results were analyzed using unpaired t-test when two groups were compared . Mann-Whitney ( MW ) test was used when a normality test failed . Analyses were also done between HD ( represented in the graphs as dashed line ) and patients after cure and no significant differences were found . The correlation analyses were performed using the Spearman's rank . Differences were considered to be statistically significant , when p≤0 . 05 .
High levels of the pro-inflammatory cytokines , IL-6 and IL-8 , and regulatory cytokine IL-10 , were found in the circulation of P . vivax-infected patients before treatment initiation , when compared to the same patients after anti-malarial therapy ( Figure 1A ) . While the cytokinemia of P . vivax-infected patients was significantly higher than individuals after cure , the absolute numbers of total leukocytes decreased with infection ( Figure 1B ) . The frequencies of lymphocytes were lower , whereas the proportions of polymorphonuclear cells and monocytes were higher in symptomatic malaria patients . Monocyte frequencies were also assessed within PBMC by flow cytometry . The frequencies of CD14+ monocytes were significantly higher in P . vivax-infected patients than in the same individuals after treatment ( Figure 1C ) . The expression of the activation marker , HLA-DR , cell adhesion molecules , CD54 , CD106 , CD31 , and chemokine receptors , CXCR3 , CCR7 , was analyzed on circulating monocytes ( Figure 2 ) . Significantly lower levels of HLA-DR were found on monocytes from P . vivax-infected patients when compared to the same patients after treatment ( Figure 2 ) . Monocytes from acute malaria patients also displayed significantly lower levels of the adhesion molecule CD31 and the chemokine receptor CCR7 ( Figure 2 ) . In contrast , significantly higher expression of the adhesion molecules CD106 ( VCAM-1 ) , CD54 ( ICAM-1 ) , and the chemokine receptor CX3CR1 , were observed on monocytes from acute malaria patients before treatment initiation ( Figure 2 ) . The expression of all these molecules on monocytes from malaria patients reached the levels found in healthy donors when analyzed 30 days after treatment . Thus , monocytes from P . vivax-infected patients exhibit a distinct activation state during acute infection . As noted above , human monocyte subsets can be distinguished by flow cytometry based on the expression of CD14 and CD16 [18] . The majority of monocytes express CD14 but not CD16 , and those expressing CD16 can be subdivided into two subpopulations , CD14+CD16+ and CD14loCD16+ cells ( Figure 3A ) . They are also slightly different in granularity as previously described and shown in Figure S1 [18] . To exclude any contaminating neutrophils , only HLA-DR+ cells were included in the analysis and CD16+ cells were included only if they also expressed HLA-DR . Higher frequencies of CD14+CD16− monocytes were found in P . vivax-infected patients when compared with patients after treatment ( Figure 3B ) . The frequencies of CD14+CD16+ and CD14loCD16+ monocytes did not differ significantly between malaria patients before and after treatment . Chemokine receptors and the adhesion molecule , LFA-1 , have been reported to be differently expressed on monocyte subsets . Previous studies described that CCR2 is expressed by both CD14+CD16− and CD14+CD16+ but not by CD14loCD16+ monocytes [18] , [27] . Similarly in P . vivax-infected patients , low levels of CCR2 were found in CD14+CD16− and CD14+CD16+ monocytes and , as expected , CCR2 was barely expressed by CD14loCD16+ cells ( Figure 3C and 3D ) . In addition , no changes in LFA-1 expression in the different monocyte subsets were observed in P . vivax patients as a result of treatment ( Figure 3C ) . Higher LFA-1 levels were observed in CD14+CD16+ , followed by CD14loCD16+ and CD14+CD16− monocytes ( Figure 3D ) . The expression of CX3CR1 along with LFA-1 has been implicated in the ability of CD14loCD16+ monocytes to crawl on the inner surface endothelium of blood vessels [18] , [28] . Importantly , P . vivax infection triggered a significantly increased expression of CX3CR1 on CD14+CD16− , CD14+CD16+ and CD14loCD16+ monocytes ( Figure 3C ) . CX3CR1 was expressed at lower levels by CD14+CD16− monocytes , when compared to the CD16+ populations ( Figure 3D ) . Lastly , changes in CCR7 expression were observed only on CD14+CD16− monocytes , with those from P . vivax infected patients expressing lower levels of CCR7 compared to monocytes from treated individuals ( Figure 3C ) . Thus , all the monocyte subpopulations from P . vivax-infected display a distinct phenotypic profile in patients undergoing acute malaria compared with the same individuals after treatment . We next FACS-sorted the CD14+CD16− , CD14+CD16+ and CD14loCD16+ monocyte subpopulations from healthy donors and P . vivax-infected patients ( Figure 4A ) . The purity of each cell population is shown in Figure 4A . Since the numbers of each circulating monocyte subset obtained from the FACS-sort was limited , we chose to assess the expression of several genes involved in inflammatory responses by nanostring [25] . The expression of 72 selected genes involved in innate immune response , cell adhesion , migration and phagocytosis was evaluated , and we found that 41 genes had their expression significantly altered in at least one of the monocyte subpopulations upon P . vivax infection ( Figure 4B ) . Differences greater than 4-fold cannot be appreciated in the heatmap , once the range from −4 . 0 to +4 . 0 fold was selected to better reveal differences in the majority of the genes induced by malarial infection in monocyte subsets . Once changes in gene expression were detected upon P . vivax infection , the expression of each of these genes was compared among the monocytes subset from P . vivax-infected patients ( Figure 4C ) . In general , the CD14+ subpopulations , i . e . , the classical and inflammatory monocytes , expressed higher levels of RNAs encoded by pro-inflammatory genes . The chemokines CCL2 and CXCL2 were highly expressed by CD14+CD16− and CD14+CD16+ subpopulations , but were expressed at lower levels by CD14loCD16+ monocytes ( Figure 4C ) . The same pattern of expression was observed for TNFR1/TNFRSF1A and ICAM-1 with CD14+CD16− and CD14+CD16+ expressing higher levels than CD14loCD16+ monocytes ( Figure 4C ) . The classical monocytes , CD14+CD16− expressed higher counts of mRNA for the receptor for IFN-gamma and for the IL-1 receptor agonist IL-1RA ( Figure 4C ) . The expression of cytokine genes also varied among monocyte subsets . Both IL6 and IL10 had higher expression in CD14+CD16+ compared to CD14+CD16− cells . CD14loCD16+ monocytes expressed higher levels of mRNA for TNF and lower levels of IL8 than the CD14+CD16− subset . CD14+CD16− as well as CD14+CD16+ monocytes expressed higher amounts of NLRP3 and CASPASE1 than patrolling monocytes . The same pattern of expression was observed for NFKB1 and NFKB1A , involved respectively with induction and regulation of cytokine expression , with CD14+CD16− and CD14+CD16+ expressing higher levels than CD14loCD16+ monocytes . REL that encodes a protein that is a member of the Rel/NFKB family was more highly expressed on CD14+CD16+ monocytes than on the other monocyte subpopulations . The expression of the costimulatory molecule CD80 was higher in CD14+CD16+ than classical and patrolling monocytes ( Figure 4C ) . Together these data indicate that CD14+CD16− and CD14+CD16+ monocytes have a more activated and inflammatory profile than patrolling monocytes during malaria . Patrolling monocytes can be distinguished from the classical and inflammatory subsets based on size and granularity [18] , but no ultrastructural analysis had been previously performed on these cell subpopulations . Electronic microscopy was performed attempting to reveal morphological changes suggestive of functional alterations . FACS-sorted CD14+CD16− , CD14+CD16+ and CD14loCD16+ monocytes from healthy donors ( n = 5 ) and P . vivax-infected patients ( n = 6 ) ( Figure 4A ) were fixed and processed for ultrastructural analysis by electron microscopy ( Figure 5A ) . CD14+CD16− and CD14+CD16+ monocytes from HD ( Figure 5A , upper panel ) and P . vivax-infected patients ( Figure 5A , lower panel ) had a larger and a higher number of mitochondria ( white arrows ) when compared to CD14loCD16+ monocytes . All monocyte subsets from P . vivax-infected patients displayed morphological features compatible with activation ( Figure 5A ) . Moreover , mitochondria from CD14+CD16+ cells from P . vivax-infected patients were significantly larger than those in the two other monocyte subsets , when mitochondria area was measured using the software ImageJ 1 . 47K ( NIH ) ( Figure 5B ) . Mitochondria area was assessed in at least six cells of each monocyte subpopulation per patient . ROS are generated in multiple compartments and by multiple enzymes in the cell and important contributions include proteins within the plasma membrane , e . g . , NADPH oxidases , and mitochondria [29] , [30] . Since mitochondria are at least in part responsible for the generation of ROS [31] , [32] , we further assessed the content of mitochondria in the monocyte subsets from P . vivax-infected patients . MitoTracker Red CMX-Ros was used for this propose . The MFI of MitoTracker Red , probe sensitive to membrane potential , was assessed in CD14+CD16− , CD14+CD16+ and CD14loCD16+ monocyte subsets by flow cytometry ( Figure 5C ) . CD14+CD16− and CD14loCD16+ monocytes similarly react with Mitotracker Red while significantly higher reactivity was found in CD14+CD16+ monocytes ( Figure 5C ) . Our data show that CD14+CD16+ monocytes have larger and more active mitochondria suggesting differential metabolic activity during P . vivax infection . The expression of p47phox and p67phox , cytosolic components of the NADPH oxidase , was also measured in monocyte subsets from malaria patients after a short-term culture with P . vivax-infected reticulocytes . Higher expression of p47phox and p67phox were found in CD14+CD16+ monocytes when comparing to their other counterparts ( Figure 5D ) . Upon activation monocytes undergo several changes , including expression of molecules involved with antigen presentation , cell adhesion and migration [33] , [34] , [35] . CD14+CD16− monocytes from patients undergoing P . vivax infection expressed lower levels of HLA-DR ( Figure 6A ) . In contrast , the expression of HLA-DR did not differ in CD14+CD16+ and CD14loCD16+ monocytes when cells from malaria patients were compared before and after treatment ( Figure 6A ) . Higher levels of HLA-DR were found in CD14+CD16+ monocytes compared to their other counterparts during P . vivax infection ( Figure 6A , B , C ) . We also observed that P . vivax infection triggered the expression of VCAM-1 and ICAM-1 in all three monocyte subpopulations ( Figure 6 ) . In contrast , a decreased expression of PECAM-1 was observed in patients experiencing malaria when compared to the same patients after treatment ( Figure 6A ) . Importantly , CD14+CD16+ monocytes from P . vivax-infected patients expressed the highest levels of ICAM-1 and PECAM-1 , when compared to CD14+CD16− and CD14loCD16+ monocytes ( Figure 6B , C ) . Taken together these results further corroborate that CD14+ monocytes , especially the CD14+CD16+ subset , are highly activated during P . vivax malaria . We used CFSE labeled P . vivax-infected reticulocytes ( Pv-Ret ) to quantify phagocytosis by different monocyte subpopulations from malaria patients before treatment initiation . CD14+CD16+ cells displayed significantly higher levels of phagocytosis of Pv-Ret than the other monocyte subsets ( Figure 7A ) . The phagocytic ability of CD14+CD16+ cells was followed by the CD14loCD16+ patrolling monocytes , which was significantly better than the CD14+CD16− monocytes ( Figure 7A ) . It is important to note that significant differences are found in the phagocytic ability of monocyte subsets when uninfected reticulocytes are purified from healthy donors and co-cultured with PBMC from patients and healthy donors . Much lower frequencies of reticulocytes containing monocytes are detected when cultures are performed with uninfected reticulocytes compared to P . vivax-infected reticulocytes ( CD14+CD16−: 3 . 54%±0 . 52% vs . 71 . 14%±11 . 11% , CD14+CD16+: 23 . 20%±2 . 18% vs . 98 . 33%±1 . 40% , CD14loCD16+: 23 . 18%±4 . 22% vs . 87 . 13%±8 . 10% ) . The same phenomenon was shown for P . falciparum [36] . Interestingly , higher levels of phagocytosis of Pv-Ret correlated with higher expression of the adhesion molecules ICAM-1 ( CD54 ) , LFA-1 ( CD11a ) and PECAM-1 ( CD31 ) by monocytes ( Figure 7B ) . As adhesion molecules , such as CD54 , expressed by cell lines interact with P . vivax-infected reticulocytes [22] , we used blocking antibodies to assess whether phagocytosis of P . vivax-infected erythrocytes was dependent on those interactions . Indeed , the phagocytosis of Pv-Ret by CD14+ cells was partially blocked by either anti-CD54 or -CD11a or -CD31 , as observed by the significantly decreased frequencies of monocytes containing Pv-Ret ( Figure 7C ) . To assess the ability of each monocyte subpopulation to kill Pv-Ret , PBMC were left in cultures for 12 h in the presence of CFSE-labeled-Pv-Ret , and the mean fluorescence intensity of CFSE were measured in each monocyte subset . The MFI of CFSE was only decreased in CD14+CD16+ monocytes when analyzed 12 h and compared to 4 h of culture ( Figure 7D ) . Both CD14+CD16− and CD14l°CD16+ displayed similar levels of Pv-Ret containing monocytes when the MFI was compare between 4 and 12 h of culture . Important to mention that no significant increase in apoptosis or changes in the proportions of monocyte subsets were detected when PBMC were cultured for 4 and 12 hours in medium or Pv-Ret ( Figure S2 ) ( Text S1 ) . To examine the mechanism of killing , we evaluated the ability of the different monocytes to produce TNF-α and ROS , which are key effector molecules made by activated monocyte/macrophages . Significantly higher frequencies of TNF-α producing cells were found among CD14+CD16+ monocytes when PBMC were cultured with Pv-Ret ( or LPS ) ( Figure 7E ) . ROS production by PBMC and monocyte subsets from P . vivax-infected patients was measured by luminescence ( RLU ) of luminol or fluorescence ( RFU ) of H2DCFDA ( Figure S4 ) . PBMC from P . vivax-infected patients produced detectable amounts of total ROS spontaneously or in response to PMA , when measured by luminol ( Figure S4A ) ( Text S1 ) . PBMC from acutely infected patients produce higher levels of ROS than the same patients after treatment . In addition , total ROS production was measured in purified monocyte subsets , but no significant differences were found among them ( Figure S4B ) . Intracellular ROS production by each monocyte subpopulation was also measured in the single-cell level by flow cytometry . Consistent with the phagocytosis and TNF-α results , CD14+CD16+ monocytes exposed to Pv-Ret generated significantly higher levels of ROS than the other monocyte subpopulations ( Figure 7F ) . We then performed experiments using rotenone and DPI ( diphenylene iodonium ) to block respectively the mitochondria complex I and NADPH oxidase plus nitric oxide synthases [37] , [38] , [39] . The inhibition of ROS production was assessed in PBMC by luminescence using luminol and in monocyte subsets using H2DCFDA by flow cytomety ( Figure S4C and S4D ) . Rotenone was able to partially inhibit ROS production by PBMC ( Figure S4C ) and by CD14+CD16+ cells cultured with parasitized reticulocytes ( Figure S4D ) . When DPI was added in the culture , it was able to abrogated ROS production by PBMC ( Figure S4C ) . Partial inhibition was observed when H2DCFDA+CD14+CD16+ cells from malaria patients were cultured with Pv-Ret and assessed by flow cytometry ( Figure S4D ) . To more specifically analyze whether mitochondrial ROS were differentially produced by monocyte subsets during malaria , MitoSox staining was performed . Corroborating with total intracellular ROS production assessed with H2DCFDA , CD14+CD16+ cells were the monocyte subset that most reacted with MitoSox ( Figure 7G ) . Taken together , our data indicate that CD14+CD16+ inflammatory monocytes play important effector activity during P . vivax infection .
Although pro-inflammatory cytokines play an important role in host resistance to Plasmodium infection , various studies reported that they may contribute to deleterious effects during malaria [40] , [41] , [42] . Thus , the pathways involved on induction of these mediators during malaria , represent checkpoints for immunological intervention to prevent poor outcome of disease . Monocytes have been described as a major source of cytokines during P . vivax infection [5] . Although infection with Plasmodium spp is known to dramatically alter monocyte differentiation [43] , the role of monocyte subsets in host resistance to infection and pathogenesis of malaria remains poorly understood . The findings reported here clearly demonstrate that both the classical ( CD14+CD16− ) and intermediate or inflammatory ( CD14+CD16+ ) monocytes are important sources of cytokines during acute P . vivax infection . Intriguingly , the CD14+CD16+ cells displayed the highest mitochondria content and activity , being an important source of ROS and were the most efficient phagocytes of P . vivax infected reticulocytes . Both in mice and human , different monocyte subsets seem to reflect developmental stages with distinct physiological roles , such as recruitment to inflammatory lesions or entry to normal tissues [17] . Consistent with our results assessing the monocyte subsets in P . vivax-infected patients , the majority of monocytes found in steady state , known as classical monocytes , express CD14 but not CD16 , and the remaining monocytes express both CD14 and CD16: CD14+CD16+ ( inflammatory ) and CD14loCD16+ ( patrolling ) monocytes [18] . This classification still gives rise to discussion . Ziegler-Heitbrock and coworkers has defined CD14loCD16+ as non-classical and CD14+CD16+ as intermediate monocytes [44] . Moreover , some studies have been analyzed monocytes based on the expression of molecules related to differentiation/activation found in macrophages [45] , [46] . Despite several of these molecules were assessed in this study many others , such as CD68 , CD163 and CD206 have been strongly correlated with different monocyte subsets [45] . As described here , both classical and inflammatory monocytes expressed the chemokine receptor CCR2 . It was previously shown in the P . chabaudi rodent model of malaria that inflammatory monocytes migrate to spleens , in CCR2 dependent manner , where they are important effector cells implicated in the control of parasite burden , likely through their phagocytic activity and release of ROS [47] . CCR2 , the chemokine receptor for CCL2 ( also known as monocyte chemotactic protein-1 ) is a marker for inflammatory monocytes . These monocyte subsets were also shown to express higher levels of mRNA encoding CCL2 when compared to the patrolling monocytes . Similarly , a previous study reported that patients infected with either P . vivax or P . falciparum have high levels of circulating CCL2 [48] . Classical and inflammatory monocytes from P . vivax-infected patients also expressed high levels of inflammatory mediators , including CXCL2 and the receptors for TNF-α , IFN-γ and IL-1 . Despite the similarities described above , CD14+CD16+ monocytes displayed the highest frequencies of TNF-α producing cells when exposed to Pv-Ret . Interestingly , CD14+CD16+ monocytes also expressed higher mRNA for IL-10 than the other monocyte subsets . It has been described that highly activated effector cells can acquire regulatory features . During Leishmania [49] , [50] , T . cruzi [51] and T . gondii [52] infection polarized Th1 cells produce IL-10 along to IFN-γ , in attempt to control immunopathology . The same has been described for monocytes . A recent article shows in the murine model of toxoplasmosis that Ly6Chi monocytes entering the gastrointestinal tract responded to commensal ligands by adopting a regulatory phenotype . For instance inflammatory monocytes became capable to control parasite burden while limiting collateral damage to tissue [53] . Indeed , plasma levels of IL-10 are lower with increased disease severity during P . vivax infection [42] . Thus , the expression of this counter regulatory cytokine may also represent an important role of CD14+CD16+ monocytes in preventing immunopathology during P . vivax malaria . Different adhesion molecules , including CD36 , ICAM-1 , VCAM-1 and PECAM-1 have been described as important receptors that bind P . falciparum infected red blood cells and influence the outcome of disease [54] , [55] . P . falciparum-infected erythrocytes are able to tether and roll on CD36 , ICAM-1 , P-selectin , and VCAM-1 in a shear-dependent fashion . In addition , CD36 has an important role in phagocytosis of P . falciparum infected cells [56] , [57] , [58] . On the other hand , ICAM-1 , but not CD36 , was implicated in the cythoadhesion of P . vivax to endothelium cells [22] . We found that CD14+CD16+CCR2+ inflammatory monocytes from P . vivax malaria patients express the highest levels of ICAM-1 , PECAM-1 , and LFA-1 . Although these adhesion molecules were originally identified as endothelium receptors for parasitized red blood cells , their expression on monocytes may favor biding and uptake of P . vivax infected reticulocytes . Indeed , our results indicate that the phagocytic activity of different monocyte subsets positively correlated with the expression of ICAM-1 , PECAM-1 and LFA-1 and blockade of each of these adhesion molecules efficiently inhibited phagocytosis of Pv-Ret . It is noteworthy that CD14+CD16+ monocytes have also been reported to expand in a group of P . falciparum-infected patients , and total monocytes from these patients were able to better control parasite growth in vitro , through antibody dependent cellular inhibition [59] , [60] . CD16 , a Fcγ receptor , has a high affinity for IgG1 and IgG3 [61] and therefore may be involved in phagocytosis of Plasmodium infected red blood cells . In contrast , experiments performed with Staphylococcus aureus or Echerichia coli showed that CD14+ monocytes have higher phagocytic activity than CD14lo monocytes . However no differences were observed between CD14+CD16+ and CD14+CD16− monocytes [62] . In our system , the expression of CD16 was not the only molecule involved in phagocytosis since the ability of Pv-Ret internalization was different between CD14+ and CD14lo , both expressing CD16 . The expression of adhesion molecules ICAM-1 , PECAM-1 and LFA-1 , though , appeared to play an important role in parasite internalization by inflammatory monocytes . In vitro studies have shown that hemozoin interferes in the upregulation of MHC class II and CD54 on monocytes after IFN-γ stimulation [63] , altering also their differentiation and maturation in dendritic cells [64] and antigen presentation [65] . Despite hemozoin is known to impair the ability of monocytes to repeat phagocytosis [66] , no pigment was found when PBMC were evaluated in our study . Moreover , impairment in phagocytosis was not observed and the monocytes still produced ROS in response to PMA , in oppose to monocytes previously exposed to parasitized red blood cells [63] . We believe that phagocytosis of P . vivax-infected reticulocytes by monocytes are taking place in the spleen , where bona fide undifferentiated monocytes reside in equivalent numbers in circulation [67] . Besides their activated phenotype , CD14+ monocytes displayed an activated morphology with larger mitochondrias . Interestingly , the CD14+CD16+ subset showed higher mitochondrial activity than the other monocyte subpopulations . Malaria infection triggers production of high levels of total ROS . Despite similar levels of total ROS are produced by different monocyte subsets , higher frequencies and levels of intracellular ROS , and higher expression of p47phox and p67phox were found in CD14+CD16+ cells compared to CD14+CD16− and CD14loCD16+ monocytes . In addition , staining with MitoSox reveals that CD14+CD16+ monocytes produce higher levels of mitochondrial ROS . ROS are important effector free radicals involved in Plasmodium killing [47] , [68] . Indeed , either blockade of mitochondrial complex I or blockage of NADPH oxidases , both responsible for ROS generation , efficiently reduces ROS levels in PBMC and CD14+CD16+ monocytes from malaria patients . Despite mitochondrial ROS have been regarded as byproducts of oxidative respiration , studies have indicated that mitochondria are recruited to vacuoles containing pathogens through an active process mediated by immune signaling [69] , [70] , [71] . Moreover , West and coworkers showed that the reduction of macrophage mitochondrial ROS results in defective bacterial killing [71] . Similarly to ROS production , higher frequencies of cells producing TNF-α , a cytokine known to trigger the respiratory burst , were found among CD14+CD16+ monocytes exposed to Pv-Ret . In a different context , patients experiencing severe malarial anemia triggered by P . falciparum display increased numbers of circulating monocytes , with significant augment in the numbers of TNF-α-producing CD14+CD16+ monocytes [72] . It is still unclear what is the cause of anemia during malaria . Despite recent reports have shown evidence demonstrating that P . vivax malaria may be associated with higher frequency and more severe anemia than P . falciparum , most of these studies were performed with children already suffering from malnutrition and hospitalized subject [73] , [74] , [75] . Only a small proportion of the P . vivax-infected patients analyzed in this study presented mild anemia ( Table S1 ) and their reticulocyte counts are according to the reference values [76] . In fact , P . vivax infection is responsible for very low frequencies of infected red blood cells . We believe that during uncomplicated vivax malaria , phagocytosis of reticulocytes by monocytes and the production of inflammatory mediators will preferentially help in the parasite control . It is important to mention that P . vivax-infected patients display elevated levels of hepatic biomarkers such as aspartate ( AST ) , alanine ( ALT ) aminotransferase and bilirubin . And despite no correlation was found between monocyte activation markers and most of the laboratory parameters , such as hematocrit and platelet counts , higher levels of AST correlated with higher expression of CCR2 and VCAM-1 on both CD14+CD16− and CD14+CD16+ monocytes and ICAM-1 on CD14+CD16− monocytes ( Figure S3 ) . These data indicate that in the attempt of controlling parasitemia , monocytes might cause inflammatory damage . Finally , we observed that CD14loCD16+ patrolling monocytes from P . vivax malaria patients did not express CCR2 but expressed high levels of LFA-1 and CX3CR1 . The latter two receptors appear to be responsible for the ability of monocytes to patrol blood vessels in vivo [18] . Consistent with this hypothesis , CD16+ monocytes display enhanced capacity to adhere to endothelial cells in vitro . This ability is partially dependent on fractalkine , the CX3CR1 ligand and CD11a , the α chain of LFA-1 [77] . Importantly , patrolling monocytes have been shown to express high levels of TLR8 and TLR9 [18] , and may also play an important role on cytokine production in response to parasite RNA and DNA . Thus , patrolling monocytes may also contribute to the early inflammatory response observed during P . vivax infection . In conclusion , our findings support the concept that highly activated monocytes are characteristic of acute malaria . P . vivax-infection leads to cytokine production by classical and inflammatory monocytes , and this response is likely to be largely responsible for many of the signs and symptoms observed in malaria sepsis . Importantly , we demonstrate for the first time that CD14+CD16+ monocytes in malaria patients exhibited greater phagocytic activity and produced higher levels of intracellular TNF-α and reactive oxygen species , indicating their important role in parasite control and host resistance to infection . Further delineation of the differential roles of monocyte subsets in P . vivax malaria could lead to identification of specific targets for therapeutic intervention in this extremely important but highly neglected parasitic disease .
|
Malaria , caused by a protozoa parasite , Plasmodium , affects more than 200 million people per year . The infection triggers an acute febrile illness , the paroxysms , occurring every 48 or 72 hours depending on the species . Plasmodium vivax , in most cases , does not cause severe malaria , but it is the most geographically widespread parasite responsible for human disease and causes substantial costs to individuals and governments . Once the parasite reaches the blood stream , they infect reticulocytes that can be destroyed by phagocytes . Our goal was to assess the importance of monocyte subsets during malaria . We found that P . vivax infection causes an increase in frequency of circulating monocytes , which were defined as classical , inflammatory , and patrolling , based on the expression of membrane molecules . Classical and inflammatory monocytes produced higher levels of pro-inflammatory cytokines and were distinguished from patrolling monocytes by displaying larger and more active mitochondria . Importantly , inflammatory monocytes were more efficient phagocytes; produced high levels of intracellular reactive oxygen species and TNF and consequently control better Plasmodium vivax infection . Hence , our results support the hypothesis that CD14+CD16+ monocytes display effector functions involved in parasite control during malaria .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"immune",
"response",
"biology",
"and",
"life",
"sciences",
"immunology"
] |
2014
|
The CD14+CD16+ Inflammatory Monocyte Subset Displays Increased Mitochondrial Activity and Effector Function During Acute Plasmodium vivax Malaria
|
An epidemiological study of leishmaniasis was performed in Amazonian areas of Ecuador since little information on the prevalent Leishmania and sand fly species responsible for the transmission is available . Of 33 clinical specimens from patients with cutaneous leishmaniasis ( CL ) , causative parasites were identified in 25 samples based on cytochrome b gene analysis . As reported previously , Leishmania ( Viannia ) guyanensis and L . ( V . ) braziliensis were among the causative agents identified . In addition , L . ( V . ) lainsoni , for which infection is reported in Brazil , Bolivia , Peru , Suriname , and French Guiana , was identified in patients with CL from geographically separate areas in the Ecuadorian Amazon , corroborating the notion that L . ( V . ) lainsoni is widely distributed in South America . Sand flies were surveyed around the area where a patient with L . ( V . ) lainsoni was suspected to have been infected . However , natural infection of sand flies by L . ( V . ) lainsoni was not detected . Further extensive vector searches are necessary to define the transmission cycle of L . ( V . ) lainsoni in Ecuador .
Leishmaniases are caused by infection with protozoan parasites of the genus Leishmania transmitted by bites of female sand flies [1 , 2] . The genus Leishmania is further divided into two subgenera , Leishmania ( Leishmania ) and Leishmania ( Viannia ) , originally distinguished by their development in the digestive tract of sand fly vectors and later confirmed by phylogenetic studies . This group of diseases is distributed worldwide , especially in tropical and subtropical areas , affecting at least 12 million people in 98 countries [2] . Approximately 20 Leishmania species are known to be pathogenic to humans , and the infecting species is the major determinant of clinical outcome [2] . Therefore , identification of the parasite species in endemic areas is important for both appropriate treatment and prognosis . In Ecuador , leishmaniasis is a major public health concern and is reported in 21 of 24 provinces of the country , in Pacific coast subtropical , Amazonian , and Andean highland areas [3] . Currently , seven Leishmania species , Leishmania ( Leishmania ) mexicana , L . ( L . ) amazonensis , L . ( L . ) major-like , Leishmania ( Viannia ) guyanensis , L . ( V . ) panamensis , L . ( V . ) braziliensis , and L . ( V . ) naiffi , have been identified as causative agents for human cutaneous ( CL ) and mucocutaneous leishmaniases ( MCL ) [4 , 5 , 6] . In Pacific coast areas and Andean areas , causative parasite species have been studied extensively; L . ( V . ) guyanensis , L . ( V . ) panamensis , L . ( V . ) braziliensis , and L . ( L . ) amazonensis in Pacific areas , and L . ( L . ) mexicana as a dominant species and L . ( L . ) major-like as a minor species in Andean areas [4 , 5 , 7 , 8 , 9 , 10 , 11] . On the other hand , in Amazonian areas , although L . ( V . ) guyanensis , L . ( V . ) braziliensis , and recently , L . ( V . ) naiffi have been identified as causative agents for CL and MCL [4 , 5 , 6 , 12 , 13 , 14] , epidemiological studies on leishmaniasis have been very limited and consequently little information is available regarding prevalent parasite species as well as epidemiological situation . Molecular biological methods are widely used for identification of Leishmania species using DNA extracted from clinical material of patients’ lesion , and is a powerful tool for epidemiological studies of leishmaniasis [15 , 16 , 17 , 18] . DNA extracted from Giemsa-stained smears obtained from patients’ skin ulcers , which are used routinely for microscopic diagnosis in the laboratory to detect parasites in the lesion , have been used also as templates for detection and identification of Leishmania DNA although the sensitivity is low because of the limitation of the DNA source [19 , 20 , 21 , 22] . Recently , to facilitate sample collection and DNA extraction processes , a Flinders Technology Associates ( FTA ) card ( Whatman ) , a filter paper that readily lyses the spotted materials and fixes nucleic acids , was used for direct sampling of patients’ material in an epidemiological study of leishmaniasis , and its usability was reported [18 , 23] . In the present study , using smear slides and FTA card-spotted samples as DNA sources , an epidemiological survey of leishmaniasis was performed in Amazonian areas where little information on the endemic Leishmania species and vector sand flies is available .
Clinical samples were collected from patients suspected of having CL who visited health centers in Cascales , Lago Agrio ( Province of Sucumbíos ) , Coca , La Joya de los Sachas , and Nuevo Rocafuerte ( Province of Orellana ) for the diagnosis and treatment of leishmaniasis ( S1 Fig ) . Tissue samples were taken by scraping the margins of active lesions on patients , spotting them onto FTA Classic Cards ( Whatman , Newton Center , MA ) and storing them at room temperature . Two-mm-diameter disks containing the sample spot were punched out from each filter paper and washed three times with FTA Purification Reagent ( Whatman ) and once with Tris-EDTA buffer . The disks were air-dried and directly subjected to PCR amplification . For the extraction of DNA from Giemsa-stained smears taken from patients’ skin ulcers that were used for diagnosis of CL , 30 μl of DNA extraction buffer [150 mM NaCl , 10 mM Tris-HCl ( pH 8 . 0 ) , 10 mM EDTA and 0 . 1% sodium dodecyl sulfate ( SDS ) ] containing 100 μg/ml of proteinase K were spotted on each smear , and the tissue sample was collected into 1 . 5 ml microtube . The sample was incubated at 37°C overnight , heated at 95°C for 5 min , and then 0 . 5 μl of each sample was directly used as a template for PCR amplification . Sand flies were captured with mouth aspirator on protected human bait , CDC light traps , and the modified Shannon light traps [24] between 18:30 and 21:00 for 14 nights on February 2015 at mountain and forest areas around patient houses in Sucumbíos Province where a patient was suspected to have acquired an infection . Female sand flies were dissected and identified to species level based mainly on the morphology of their spermathecae [25] . Sand flies were examined under light microscopy for natural flagellate infections and positive samples were fixed individually in absolute ethanol . Ethanol-fixed specimens were individually lysed in 50 μl of DNA extraction buffer with proteinase K , and 0 . 5 μl of the extract was directly used as PCR templates [6 , 26 , 27 , 28] . Leishmania species were identified by cytochrome b ( cyt b ) gene sequence analysis [18 , 23] . PCR amplification with a pair of specific primers , L . cyt-AS ( 5'-GCGGAGAGRARGAAAAGGC-3' ) and L . cyt-AR ( 5'-CCACTCATAAATATACTATA-3' ) , was performed with 30 cycles of denaturation ( 95°C , 1 min ) , annealing ( 55°C , 1 min ) and polymerization ( 72°C , 1 min ) using Ampdirect Plus reagent ( Shimadzu Biotech , Tsukuba , Japan ) . A portion of the PCR product was reamplified with L . cyt-S ( 5'-GGTGTAGGTTTTAGTYTAGG-3' ) and L . cyt-R ( 5'-CTACAATAAACAAATCATAATATRCAATT-3' ) . For some samples , Leishmania species were further identified by heat-shock protein 70 ( hsp70 ) gene sequence analysis [29] . PCR amplification with a pair of specific primers , HSP70sen ( 5'-GACGGTGCCTGCCTACTTCAA-3' ) and HSP70ant ( 5'- CCGCCCATGCTCTGGTACATC-3' ) , was performed with 40 cycles of denaturation ( 95°C , 1 min ) , annealing ( 55°C , 1 min ) and polymerization ( 72°C , 1 min ) using Ampdirect Plus reagent ( Shimadzu Biotech , Tsukuba , Japan ) . The products were cloned into the pGEM-T Easy Vector System ( Promega , Madison , WI ) and sequences were determined by the dideoxy chain termination method using a BigDye Terminator v3 . 1 Cycle Sequencing Kit ( Applied Biosystems , Foster City , CA ) . The Leishmania cyt b gene sequences were aligned with CLUSTAL W software [30] and examined using the program MEGA ( Molecular Evolutionary Genetics Analysis ) version 6 [31] . Phylogenetic trees were constructed by the neighbor-joining method with the distance algorithms available in the MEGA package [30] . Bootstrap values were determined with 1 , 000 replicates of the data sets . The database for phylogenetic analyses consisted of cyt b gene sequences from L . ( L . ) infantum ( GenBank accession number: AB095958 ) , L . ( L . ) donovani ( AB095957 ) , L . ( L . ) major ( AB095961 ) , L . ( L . ) tropica ( AB095960 ) , L . ( L . ) amazonensis ( AB095964 ) , L . ( L . ) mexicana ( AB095963 ) , L . ( V . ) panamensis ( AB095968 ) , L . ( V . ) guyanensis ( AB095969 ) , L . ( V . ) braziliensis ( AB095966 ) , L . ( V . ) lainsoni ( AB433280 ) , L . ( V . ) naiffi ( AB433279 ) and L . ( V . ) shawi ( AB433281 ) . The collection was performed by local physicians and well-trained laboratory technicians with the approval of the research ethics committee of the Graduate School of Veterinary Medicine , Hokkaido University ( license number: vet26-4 ) . Informed consent was obtained from the adult subjects and from the children’s parents or guardians , prior to collection of diagnostic materials at each health center of the Ecuadorian Ministry of Health ( Provinces of Sucumbíos and Orellana ) . Signed consent was obtained after the explanation of the process of diagnosis and Leishmania species analysis at the time of routine diagnosis carried out at rural health centers , following the guidelines of the Ethics Committee of the Ministry of Health , Ecuador . The subjects studied were volunteers in routine diagnosing/screening and treatment programs promoted by the Ministry . All routine laboratory examinations were carried out free of charge , and treatment with specific drug ( Glucantime ) was also offered free of charge at each health center of the Ministry .
Of 33 clinical samples , nine samples from patients in Sucumbíos Province and 24 from patients in Orellana Province were obtained ( Table 1 ) . Of these , 6 and 14 samples from Sucumbíos and Orellana patients , respectively , were Giemsa-stained smears used for routine diagnosis of CL , and others ( 3 and 10 samples , respectively ) which were specimens spotted on FTA cards ( Table 1 ) . Patients had one to four skin lesions on their face , arms or/and legs with a diameter of 1 to 3 cm , typical of those observed on patients with CL in the area . The leishmanial cyt b gene was successfully amplified from two smears and three FTA card samples from Sucumbíos , and seven smear and nine FTA card samples from Orellana ( Table 1 ) , and sequences were determined . The cyt b gene sequences from two Sucumbíos and 13 Orellana patients had a greater degree of homology with those of L . ( V . ) guyanensis ( 98 . 1–99 . 8% ) , and cyt b gene sequences from two Sucumbíos and two Orellana patients had a greater degree of homology with those of L . ( V . ) braziliensis ( 98 . 6–100% ) . In addition , cyt b sequences of one specimen each from Sucumbíos and Orellana patients who have never traveled abroad showed highest homology with those of L . ( V . ) lainsoni ( 99 . 0% and 99 . 1% ) ( S2 Fig ) . These results were supported by a phylogenetic analysis showing that parasites from 15 patients were located in the clade of L . ( V . ) guyanensis , four specimens were in the L . ( V . ) braziliensis clade , and two samples were in the L . ( V . ) lainsoni clade ( Fig 1 ) , indicating that the causative parasite species were L . ( V . ) guyanensis , L . ( V . ) braziliensis , and L . ( V . ) lainsoni , respectively . A sample , 13-8EC7 , from Sucumbíos , in which causative parasite was identified as L . ( V . ) lainsoni , was further subjected to hsp70 gene sequence analysis . The hsp70 gene sequence of 13-8EC7 had a greater degree of homology with those of L . ( V . ) lainsoni ( 99 . 3% ) , and a phylogenetic analysis supported the result ( S3 Fig ) . Nucleotide sequence data reported are available in the DDBJ/EMBL/GenBank databases under the accession numbers LC055616 , LC055617 , and LC055622-LC055638 . A total of 1 , 104 female sand flies were captured and dissected . Of these , 732 were captured on protected human bait , and 372 flies with light traps ( CDC light trap and the modified Shannon trap ) ( Table 2 ) . Four species , Lu . yuilli yuilli ( 71 . 7% ) , Lu . tortura ( 12 . 3% ) , Lu . davisi ( 7 . 3% ) , and Lu . napoensis ( 5 . 0% ) accounted for 96 . 3% of the sand flies ( Table 2 ) . Among these , Lu . napoensis was captured only by light traps . In addition to these four species , Lu . sherlocki , Lu . trapidoi , Lu . gomezi , Lu marinkelli , Lu dysponeta , Lu camposi , Lu . robusta , Lu hirusta hirusta , Lu micropyga , and five unidentified species were captured ( Table 2 ) . Natural flagellate infections were observed in the hindguts of 14 Lu . yuilli yuilli ( 1 . 8% ) , one Lu . davisi ( 1 . 2% ) , and of the only collected specimen of Lu . camposi ( Table 2 ) . Genomic DNAs were extracted from dissected sand flies infected with flagellates , and parasite cyt b genes were amplified . The cyt b gene fragments were successfully obtained from nine of the 14 positive Lu . yuilli yuilli and from Lu . davisi . The unsuccessful amplification of parasite cyt b genes from the other five Lu . yuilli yuilli and Lu . camposi was attributed to the very small number of parasites present in the gut . The cyt b gene sequences of parasites from the 10 flagellate-positive sand flies were analyzed and compared to those of related parasite species . The cyt b gene sequences from the nine Lu . yuilli yuilli showed only 86 . 6–88 . 6% homology with those of the Leishmania species , and 99 . 1–99 . 8% homology with those of Endotrypanum species , flagellate parasites of non-human animals transmitted by sand flies [32 , 33] , indicating that the flagellate infections in Lu . yuilli yuilli were Endotrypanum . On the other hand , the sequences of the parasite from Lu . davisi had only 81 . 2–82 . 3% homology with those of Leishmania and Endotrypanum species , indicating that the flagellate is neither Leishmania nor Endotrypanum . These results were supported by a phylogenetic analysis showing that nine flagellates from Lu . yuilli yuilli were located in the clade of Endotrypanum species , while the one from Lu . davisi was distant from Leishmania and Endotrypanum species ( Fig 2 ) . We suspect that the flagellates observed in Lu . davisi may belong to the genus Trypanosoma since some sand fly species are reported to transmit Trypanosoma species [28 , 34 , 35] . The vector species of L . ( V . ) lainsoni in Ecuador could not be identified in this study .
An epidemiological study of leishmaniasis was conducted in Amazonian areas of Ecuador where very little information is available on prevalent parasite species or sand fly species associated with their transmission . In addition to L . ( V . ) guyanensis and L . ( V . ) braziliensis infections , the first cases of CL caused by L . ( V . ) lainsoni infection in Ecuador were identified . A search for the vector sand fly in the area where one L . ( V . ) lainsoni infected patient presumably contracted the disease produced no positive flies , thus the vector remains unknown . Leishmaniasis is widely distributed in Pacific coast subtropical climate areas , Andean highland areas , and Amazonian tropical areas in Ecuador [4 , 5] . To date , most clinical and parasitological studies of leishmaniasis have been reported from Pacific coast and Andean highland areas [4 , 5 , 7 , 8 , 9 , 10 , 11] , and little information is available on endemic parasite and sand fly species in Amazonian areas , mainly because of difficulty in gaining access [4 , 5 , 6 , 12 , 13 , 14] . In this study , Giemsa-stained smears used for routine diagnosis of CL were utilized as a DNA source in addition to FTA card-collected samples . The detection ratio was lower in DNA samples from smear slides when compared to FTA card collection ( 2/6 vs . 3/3 in Sucumbíos and 7/14 vs . 9/10 in Orellana ) , reflecting the amount of DNA recoverable and the condition of the smear slides . The thin smear samples were methanol-fixed , Giemsa-stained , and then examined under oil immersion . Thus , part of specimens and DNA may be lost and damaged when the immersion oil was wiped from the slide . Nevertheless , stored smear slides , which will not be used for any further purpose than diagnosis by microscopic examination , can be useful for identification of Leishmania species in endemic areas where sample collection for epidemiological study is difficult . In this study , L . ( V . ) guyanensis and L . ( V . ) braziliensis were identified as causative agents in Amazonian areas as reported previously [4 , 5 , 6 , 12 , 13 , 14] . In addition , L . ( V . ) lainsoni is implicated as a causative agent of CL in Ecuador for the first time . Leishmania ( V . ) lainsoni was originally identified from patients in Brazilian Amazon [36] . It causes CL with characteristic lesions similar to those caused by other Leishmania ( Viannia ) species: small ulcers or small self-limiting nodules [37] . The parasite was subsequently identified from patients in Sub-Andean and Amazonian areas of Peru [18 , 38 , 39 , 40] , in subtropical climate areas and Sub Andean areas of Bolivia [41 , 42] , Suriname [43] , and French Giana [44] . Two L . ( V . ) lainsoni-infected patients were found during this study . One was from the northern part of Ecuadorian Amazon ( Sucumbios ) near the border with Colombia , and the other was from the eastern frontier with Peru ( Orellana ) , more than 200 km away . These findings suggest that L . ( V . ) laisoni is widely distributed in South America . Natural infections with L . ( V . ) lainsoni in sand flies have been detected in Lu . ubiquitalis in Brazil [45] , Lu . nuneztovari anglesi in Bolivia [42] , and Lu . auraensis in Peru [46] . Microscopic examinations of sand flies collected in this study revealed no natural infection with L . ( V . ) lainsoni , and none of the above three Lutzomyia species were collected . Of the three dominant human-biting species in our collections , infection with Endotrypanum species , flagellate parasites of sloths , which are non-pathogenic to humans [32 , 33] , were detected in the most dominant species , Lu . yuilli yuilli , as reported in another Amazonian area of Ecuador [6] . Although natural infection was not detected in this study , the next dominant species , Lu . tortura , was already incriminated as a vector species of L . ( V . ) naiffi in Ecuadorian Amazon [6 , 13] . Furthermore , Lu . davisi was reported to transmit L . ( V . ) naiffi in Brazil [47] and L . ( V . ) braziliensis in Peru [46] . In this study , natural infection of Lu . davisi by flagellates was observed . However , the parasite was neither Leishmania nor Endotrypanum . To identify the vector of L . ( V . ) lainsoni , further research is necessary to understand the natural transmission cycle of this parasite in Ecuador . Our finding of the first human cases of CL caused by L . ( V . ) lainsoni infection in two separate areas of Ecuador suggests that this parasite is widely distributed in South America . Extensive countrywide surveillance is necessary to gain a proper understand of the status of L . ( V . ) lainsoni , as well as the sand flies responsible for its transmission in Ecuador .
|
In Ecuador , leishmaniasis is endemic in Pacific coast subtropical , Amazonian , and Andean highland areas . Seven Leishmania species , Leishmania ( Leishmania ) mexicana , L . ( L . ) amazonensis , L . ( L . ) major-like , Leishmania ( Viannia ) guyanensis , L . ( V . ) panamensis , L . ( V . ) braziliensis , and L . ( V . ) naiffi , are reported to be associated with human cutaneous ( CL ) and mucocutaneous leishmaniases ( MCL ) . Causative parasites have been studied extensively in Pacific coast and Andean areas; however , information such as prevalent Leishmania species and their vector sand fly species is very sparse in Amazonian areas . Giemsa-stained smears taken from patients’ skin ulcers and used for routine diagnosis of CL and Flinders Technology Associates ( FTA ) card-spotted samples were utilized as DNA sources , and causative parasites were identified on the basis of cytochrome b gene analysis . Causative parasites in 25 samples were successfully identified , and , in addition to previously reported species , L . ( V . ) guyanensis and L . ( V . ) braziliensis , L . ( V . ) lainsoni was identified from two patients living in different areas situated more than 200 km apart . Sand flies were examined in areas where one of the L . ( V . ) lainsoni infected patient was suspected to have been infected . Although 1 , 104 female sand flies were dissected and examined for species identification and detection of natural infection with flagellates in the gut , human-infective Leishmania species including L . ( V . ) lainsoni were not detected . Further extensive investigation of sand fly fauna is necessary to incriminate the vector of this parasite in Ecuador .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"sequencing",
"techniques",
"medicine",
"and",
"health",
"sciences",
"ecuador",
"tropical",
"diseases",
"geographical",
"locations",
"sand",
"flies",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"protozoans",
"leishmania",
"phylogenetic",
"analysis",
"neglected",
"tropical",
"diseases",
"molecular",
"biology",
"techniques",
"insect",
"vectors",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"infectious",
"diseases",
"zoonoses",
"south",
"america",
"epidemiology",
"biological",
"databases",
"protozoan",
"infections",
"molecular",
"biology",
"disease",
"vectors",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"people",
"and",
"places",
"sequence",
"databases",
"database",
"and",
"informatics",
"methods",
"leishmaniasis",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2016
|
First Human Cases of Leishmania (Viannia) lainsoni Infection and a Search for the Vector Sand Flies in Ecuador
|
Yersinia pseudotuberculosis binds to β1 integrin receptors , and uses the type III secretion proteins YopB and YopD to introduce pores and to translocate Yop effectors directly into host cells . Y . pseudotuberculosis lacking effectors that inhibit Rho GTPases , YopE and YopT , have high pore forming activity . Here , we present evidence that Y . pseudotuberculosis selectively modulates Rho activity to induce cellular changes that control pore formation and effector translocation . Inhibition of actin polymerization decreased pore formation and YopE translocation in HeLa cells infected with Y . pseudotuberculosis . Inactivation of Rho , Rac , and Cdc42 by treatment with Clostridium difficile toxin B inhibited pore formation and YopE translocation in infected HeLa cells . Expression of a dominant negative form of Rac did not reduce the uptake of membrane impermeable dyes in HeLa cells infected with a pore forming strain YopEHJT− . Similarly , the Rac inhibitor NSC23766 did not decrease pore formation or translocation , although it efficiently hindered Rac-dependent bacterial uptake . In contrast , C . botulinum C3 potently reduced pore formation and translocation , implicating Rho A , B , and/or C in the control of the Yop delivery . An invasin mutant ( Y . pseudotuberculosis invD911E ) that binds to β1 integrins , but inefficiently transduces signals through the receptors , was defective for YopE translocation . Interfering with the β1 integrin signaling pathway , by inhibiting Src kinase activity , negatively affected YopE translocation . Additionally , Y . pseudotuberculosis infection activated Rho by a mechanism that was dependent on YopB and on high affinity bacteria interaction with β1 integrin receptors . We propose that Rho activation , mediated by signals triggered by the YopB/YopD translocon and from engagement of β1 integrin receptors , stimulates actin polymerization and activates the translocation process , and that once the Yops are translocated , the action of YopE or YopT terminate delivery of Yops and prevents pore formation .
A great spectrum of Gram-negative bacteria depends on a specialized secretion mechanism to establish a successful infection in the host . This machinery is known as the type III secretion system ( TTSS ) , and is present in organisms that are pathogenic for animals or plants , as well as in symbiotic bacteria [1] . In pathogenic Yersinia species , a TTSS is encoded in a large virulence plasmid , and is required for counteracting innate and adaptive host immune defenses [2] . This is accomplished by injection of six effector proteins ( YopE , YopT , YopH , YopJ , YopO , YopM ) that target different host cell signaling molecules . This injection mechanism is known as Yop translocation . Two effectors relevant to this work are YopE and YopT , which target a family of Rho GTPases that control a variety of cellular functions , including regulation of the actin cytoskeleton . In turn , the activity of the Rho GTPases is tightly controlled by a number of regulators . Guanine nucleotide exchange factors ( GEFs ) induce activation of GTPases by inducing GDP/GTP exchange . GTPase accelerating proteins ( GAPs ) inactivate Rho GTPases by stimulating GTP hydrolysis . Active Rho proteins are mostly associated with cellular membranes by means of a post-translational lipid modification ( prenylation ) [3] . YopE inhibits RhoGTPases by acting as a GAP for RhoA , Rac1 , or Cdc42 [4 , 5] . YopT inhibits preferably RhoA , by cleaving the isoprenyl group and removing the GTPase from the membrane [6] . Although the mechanism of translocation is not completely understood , it is thought that effectors are delivered from the bacterial cytoplasm to the outer membrane through a secretion conduit . In turn , this channel is connected to a needle–like structure that transports the effectors directly into the host cell's cytoplasm . Apart from the proteins that form the needle , three translocator proteins ( YopB , YopD and LcrV ) are required for the delivery of toxins into the host cell . YopB and YopD are thought to form a translocation channel at the plasma membrane [7–9] . Two recent report show that LcrV is located at the tip of the needle [10] , and that it may act as an assembly platform for YopB and YopD prior to their insertion in the membrane [11] . Activation upon contact of the bacteria with the host cell is one of the hallmarks of the TTSS . Adhesion of Yersinia to host cells is mediated by surface proteins , such as invasin or YadA binding to β1 integrin host cell receptors , or by pH6 antigen interacting with glycosphingolipids [12 , 13] . High affinity interaction of β1 integrin receptor with invasin , or YadA ( via fibronectin ) , stimulates a signal transduction pathway that involves activation of Src protein tyrosine kinase , tyrosine phosphorylation of focal adhesion proteins , such as FAK and Cas , and downstream activation of Rac1 and PI3-K [12 , 14 , 15] . Stimulation of this pathway results in bacterial internalization . We have previously shown that infection of epithelial cells with Y . pseudotuberculosis lacking YopE , YopT , YopJ and YopH elicits a proinflammatory signaling response that requires YopB but is independent of YopD , suggesting that this signaling event can occur in the absence of a translocation channel [16] . This proinflammatory response , characterized by activation of MAP kinases and NFκB , and production of IL-8 , is blocked by the Rho GTPase inhibitory action of YopE , and to a lesser extent YopT [17] . It is therefore possible that YopB elicits activation of a signaling pathway involving Rho GTPases . Although a translocation channel composed of YopB and YopD is thought to insert into the host cell membrane , the integrity of the plasma membrane remains intact during infection with wild type Yersinia . However , infection with Yersinia mutant strains that do not produce YopE and YopT results in loss of membrane integrity , a process known as pore formation [7 , 18] . Interestingly , yopE , yopT mutants also induce the polymerization of an actin ring at the site of the interaction with the host cell , but the link between these “actin halos” and pore formation is not known . How YopE or YopT prevent pore formation is not fully understood , and is a controversial issue [19] . We have found that , catalytically inactive forms of YopE or YopT ( [18] , unpublished data ) were not able to prevent pore formation , analyzed by uptake of impermeable dyes ( EtdBr ) or release of lactate dehydrogenase ( LDH ) . Expression of constitutively active forms of RhoA or Rac1 prior to infection , rescued the pore forming activity of bacteria expressing YopE or YopT [18] . In addition , infection carried out in the presence of actin polymerization inhibitors dramatically reduced pore formation . Based on these results we concluded that insertion of the YopB/D translocation channel results in Rho GTPases activation , actin polymerization , and pore formation [18] . Here , we present evidence that not only pore formation but most importantly , translocation is controlled by Rho activity and actin polymerization . We also found that high affinity interaction between YadA or invasin with β1 integrin receptors is crucial for efficient translocation of Yops . Thus , we hypothesize that YopB/D signaling , in cooperation with β1 integrin signaling , activates Rho to induce changes in the host cell cytoskeleton that control the translocation process .
Macrophages infected with Salmonella or Shigella species undergo a caspase-1-dependent form of cell death termed pyroptosis [20] . This death mechanism is proinflammatory , and requires Yersinia YopB homologues SipB and IpaB , from Salmonella and Shigella , respectively . A recent report shows that pyroptosis is caused by caspase-1-dependent pore formation and consequent osmotic lysis [21] . Pore formation is usually determined by the incorporation or release of membrane impermeable dyes , such as EtdBr and BCECF , respectively , by the infected cells [7 , 8 , 22] . Because pore formation is followed by osmotic lysis , an indirect method to determine pore formation involves measuring the release of the cytoplasmic enzyme lactate dehydrogenase ( LDH ) in supernatants of cultured cells [22] . In Yersinia-infected macrophages , caspase-1-mediated maturation and release of the proinflammatory cytokine interleukin 1β can be inhibited by YopE and YopT [23] . Because the inhibitory action of YopE and YopT on the Rho GTPases also blocks pore formation [18] , we investigated whether YopB/D-mediated cell lysis in HeLa cells is a result of caspase-1 mediated cell death . We used Ac-YVAD-cmk ( YVAD ) , a permeable peptide that specifically inhibits caspase-1 , irreversibly . HeLa cells treated for 1 h with 50 μM or 100 μM of YVAD , or control untreated cells , were infected with pore forming strain yopEHJ ( YP27 ) , and the corresponding pore forming-deficient strain that lacks YopB ( yopEHJB , YP29 ) . The uptake of the impermeable dye ethidium homodimer-2 ( EthD2 ) and the amount of LDH released in the supernatant of infected cells was tested 3 hours after infection . YVAD did not prevent LDH release ( Figure 1A ) or penetration of the dye ( not shown ) in cells infected with YP27 . On the other hand , YVAD treatment dramatically inhibited YP27-induced IL-1β production in J774 . 1A macrophage-like cells ( Figure S1 ) , indicating that 100μM YVAD efficiently inhibits caspase-1 mediated processes . These data support the hypothesis that YopB/D-mediated loss of membrane integrity in epithelial cells does not require caspase-1 activation . Salmonella–induced pyroptosis is also inhibited by 5 mM glycine [20] . We investigated if YopB/D-induced loss of membrane integrity could be inhibited by treatment with 5 mM glycine through out the infection . As shown in Figure 1B , glycine had no effect on the amount of LDH released by YP27-infected cells . This result further suggests that in HeLa cells YopB/D-mediated LDH release occurs by a process different from pyroptosis . We therefore consider that , in our experimental system , pore formation is linked to the translocation process . We have previously found that pore formation is prevented by the catalytic activity of two Rho GTPase-inhibiting effectors , YopE and YopT [18] . To test whether inactivation of small GTPases inhibits pore formation , we incubated cells for 2 h in the presence or absence of 40ng/ml of Clostridium difficile toxin B ( ToxB ) , an ADP-ribosylating protein that powerfully inhibits Rho , Rac and Cdc42 . ToxB treatment strongly reduced the uptake of ethidium homodimer-2 ( EthD-2 ) by cells infected with pore forming strain yopEHJ ( YP27 ) ( Figure 2A ) . Rho GTPase downregulation by ToxB also inhibited LDH release ( Figure 2B ) . Thus supernatants of YP27-infected cells treated with ToxB released levels of LDH comparable to those of cells infected with the pore-forming-deficient strain yopEHJB ( YP29 ) . These data suggest that YopB/D-mediated pore formation requires activation of Rho GTPases . We have previously observed that a catalytically inactive form of YopE ( YopER144A ) is translocated at higher levels than wild type YopE [4 , 18] . Aili et al . have also reported this phenomenon recently; they showed that several YopE mutants defective for GAP activity are hypertranslocated [24 , 25] . Interestingly , Wong and Isberg [26] observed that overexpression of YopT inhibits YopE translocation . Altogether , these observations suggest a possible role of GTPase activation in controlling the translocation process . To study this hypothesis we tested the action of ToxB on YopE translocation using the Triton X-100 solubility assay described in Material and Methods . Pretreatment of HeLa cells with ToxB reduced the amount of YopE translocated by wild type strain YP126 by 60% ( Figure 2C ) . As expected , only background levels of YopE were detected in the soluble fraction of cells infected with the translocation deficient YopB− mutant , YP18 . The inhibitory effect of ToxB on pore formation and translocation is not likely to be a consequence of an impairment of the bacteria-host cell interaction , because the number of cell-associated bacteria did not vary with ToxB treatment ( Figure S2 ) . This led us to conclude that Yop translocation is strongly influenced by the level of Rho , Rac or Cdc42 activation . In previous experiments , we have shown that actin polymerization inhibitors , cytochalasin D ( CD ) and latrunculin B , inhibit pore formation [18] . Here we confirmed the effect of CD on pore formation ( Figures 2A and 2B ) , and determined whether host actin polymerization plays a role in Yop translocation during Yersinia infection . As shown in Figure 2C , CD treatment greatly reduced the amount of translocated YopE , this inhibitory effect being comparable to the ToxB treatment . Adhesion assays showed that CD does not affect the number of cell-associated bacteria greatly ( not shown ) . These observations appear to indicate that actin polymerization is not only required for pore formation , as we had shown previously , but it also controls Yop translocation . Y . pseudotuberculosis internalization into epithelial cells requires a signaling cascade that results from the binding of invasin or YadA to β1 integrin receptors . Bacterial uptake requires small Rho GTPases activation and actin polymerization . Thus internalization , like pore formation and translocation , is inhibited by the GAP activity of YopE , and by treatment with cytochalasin D [4 , 18] . With this in mind , we investigated whether invasin or YadA-mediated adhesion to β1 integrin receptors is required for efficient pore formation and translocation . We created a yopEHJ , yadA , inv mutant strain , designated YP50 , and the corresponding YopB-deficient mutant YP51 ( Table 1 ) . To provide a means of adhesion , a pAY66 plasmid , constitutively expressing pH6 antigen ( Table 1 ) , was inserted into YP50 and YP51 . The pH6 antigen is a fimbrial adhesin that can mediate adhesion of Yersina to epithelial cells but does not induce bacterial uptake [27] . The defect in internalization of YP50/pAY66 and YP51/pAY66 was confirmed by immunofluorescence ( not shown , see below ) . To corroborate that pH6 ag can substitute invasin or YadA for adherence , we evaluated the binding ability of the YP50/pAY66 strain after one hour infection by immunofluorescence . We found that YP50/pAY66 adhered to HeLa cells at levels similar to yopEHJ ( YP27 ) expressing invasin or YadA ( not shown ) . YP50/pAY66 strain was compared to the YP27 strain for the ability to induce pore formation . Surprisingly , YP50/pAY66 caused lower levels of LDH release than YP27 ( Figure 3A ) , and was defective for promoting uptake of EthD-2 by infected HeLa cells ( not shown ) . As expected , infection with the corresponding yopB mutant YP51/pAY66 resulted in even lower levels of LDH release . Ectopic expression of YadA in the YP50 strain rescued LDH release , indicating that interaction with β1 integrin receptors is critical for pore formation . To rule out that the impairment of the inv , yadA , pH6 antigen-expressing mutant to cause pore formation was due to a defective activation of the TTSS , we tested the ability of YP50/pAY66 to induce IL-8 production , NFκB activation , and ERK phosphorylation . We have previously found that the ability to stimulate these pro-inflammatory signals requires YopB but is independent of pore formation [16] . As shown in Figure 3B , after 5 hours infection , IL-8 production was not considerably reduced by the absence of invasin or YadA . Similarly , YopB-dependent activation of NFκB and ERK , measured at 1 hour-post infection , did not require invasin or YadA ( Figure S3 ) , suggesting that YopB is able to stimulate cell responses whether adhesion is provided by invasin/YadA or by pH6 antigen . Collectively , these results indicate that interaction of the bacteria with β1 integrin receptors is required to stimulate pore formation . To investigate whether engagement of β1 integrin receptors is also needed for the translocation process , we tested the ability of a yadA , inv mutant to translocate YopE . To this end , we replaced the mutated yopE by the wild type yopE gene in YP50/pAY66 , creating YP54/pAY66 ( Table 1 ) . In line with its reduced ability to cause pore formation , the inv , yadA , pH6 antigen-expressing mutant translocated undetectable levels of YopE ( Figure 3C ) . Consequent with these findings , YP54/pAY66 induced cell rounding at a much slower rate than the wild type YP126 ( Figure 3D , compare YP126 and YP54/pAY66 after 30 min infection ) . Efficient YopE translocation was restored when YadA was expressed in YP54 ( Figure 3C ) . This suggests that the interaction of Y . pseudotuberculosis with β1 integrin receptors is required for an effective translocation process . As invasin and YadA promote both binding to β1 integrins and stimulation of signaling by this receptor , we used a mutant that is competent for binding to β1 integrins but defective in signaling , to establish which activity was important for pore formation and translocation . A single amino acid substitution , D911E , in the invasin protein retains binding to host cells , but results in low affinity interaction with β1 integrins , poor receptor clustering , and a consequent defect in signaling and internalization [28] . Thus , we assessed the ability of YP50invD911E and YP54invD911E to induce pore formation and to mediate YopE translocation , respectively . Although infection with YP50invD911E resulted in robust IL-8 production ( Figure 3B ) , the levels of LDH release by cells infected with YP50invD911E were as low as those cells infected with the strains that adhere via pH6 antigen ( Figure 3A ) . Similarly , YP54invD911E was impaired in YopE translocation ( Figures 3C and 3D ) . We ruled out that the defect in translocation was a consequence of fewer YP54invD911E bacteria binding to Hela cells . Thus , immunofluorescence analyses after 1-hour infection revealed that YP50invD911E infected cells had a mean of 16 . 6 associated bacteria/cell , only slightly lower than the invasin-expressing strain ( 19 . 7 bacteria/cell , Figure S4A ) . Moreover , a two-fold increase in the multiplicity of infection of YP54/pAY66 and YP54invD911E did not result in higher levels of YopE translocation ( Figure S4B ) . We conclude that efficient translocation and pore formation involve high affinity binding to β1 integrin receptors . To examine the binding characteristics of the inv/yadA mutant we performed transmission electron microscopy in thin section of infected HeLa cells . As expected , yopEHJ ( YP27 ) bacteria were either internalized , or were in the process of being engulfed , and tightly attached to the host cells ( Figure S5A ) . On the other hand , yopEHJ , yadA , inv/psaABC ( YP50/pAY66 ) were almost exclusively extracellular and seemed to bind more loosely ( Figure S5B ) . Adhesion mediated by invD911E differed from that imparted by wild type invasin ( Figure S5A and S5C ) . This suggests that lack of high affinity binding to β1 integrin receptors not only impairs β1 integrin signaling , but might also affect the way the bacteria interacts with the host cell . Stimulation of signaling through β1 integrins receptor by invasin and YadA involves tyrosine phosphorylation of a series of signaling proteins . Src is a key signal-transducing protein kinase in the β1 signaling pathway leading to internalization . To determine if Src activation plays a role in Yop translocation , we tested the effect of a selective inhibitor of Src family kinases , PP2 , on infected cells . Pre-treatment of cells for 1 hour with 10μM PP2 efficiently inhibited β1 integrin signaling pathway leading to bacterial internalization without decreasing bacterial adherence ( not shown ) . Interestingly , pore formation and YopE translocation were also impaired in PP2-treated cells ( Figures 4A and B ) . These data indicate that Src activation stimulates translocation , and point toward a role of β1 integrin signaling in the Yop translocation . Invasin triggered-Rac1 signaling pathways downstream of Tyr phosphorylation are essential for Yersinia uptake [15] . We made use of a specific Rac1 inhibitor to determine whether β1 integrin–mediated internalization was required for efficient pore formation and translocation . NSC23766 is a small chemical compound reported to specifically block the binding between Rac1 and its exclusive GEFs [29] . We tested the effect of the Rac1 inhibitor by pre-treating HeLa cells for 6h with 100μM of NSC23766 in DMEM with 5% serum . As expected , bacterial uptake was impaired by treatment with the Rac inhibitor , with the number of yopEHJ ( YP27 ) bacteria internalized by NSC23766-treated cells being comparable to that of the uptake-deficient yopEHJ , yadA , inv ( YP50/pAY66 ) strain ( Figure 5A ) . NSC23766 treatment was also found to inhibit formation of phagosomes , as the number of actin cups was reduced more than 5 fold in the presence of the inhibitor ( Figure S6 ) . We further excluded any effect of NSC23766 treatment on the number of cell-associated bacteria by immunofluorescence ( not shown ) . Transmission electron microscopy of thin sections also confirmed that NSC23766 inhibited bacterial uptake by , but not association to HeLa cells ( Figure S5D ) . Importantly , treatment with NSC23766 did not reduce pore formation or Yop translocation ( Figures 5B and 5C , respectively ) . These results indicate that bacterial internalization is not required for pore formation or translocation . To validate our findings using the Rac1 inhibitor , we expressed a dominant negative form of Rac1 in Hela cells . We transfected cells with a eukaryotic expression plasmid coding for a T7 tagged-RacN17 ( pCGTRacN17 ) and we evaluated whether pore formation was impaired in transfected cells . Overexpression of Rac1N17 ( green cells ) did not prevent pore formation as shown by the uptake of the impermeable dye EthD-2 ( Figures S7A and S7B ) . Altogether , these data provide evidence indicating that neither bacterial internalization , nor Rac1 activation , play a major role in the processes that govern pore formation and Yop translocation . C3 is an ADP-ribosylating protein of Clostridium botulinum that specifically inhibits Rho A , B and C . A recombinant cell-permeable form of C3 toxin ( TAT-C3 ) was produced in E . coli and purified as described in Material and Methods . Four hours before infection , HeLa cells were treated with 10 , 20 , and 40μg/ml of TAT-C3 in serum-free medium , or with serum-free medium alone . C3 has been previously shown to increase Y . pseudotuberculosis uptake in COS-1 cells [30]; in our experimental model , pretreatment of cells with 20μg/ml TAT-C3 did not affect bacterial adhesion or internalization considerably ( Figures S8A and S8B , respectively ) . Interestingly , TAT-C3 treatment of cells infected with the pore forming strain yopEHJ ( YP27 ) inhibited LDH release in a dose-dependent manner ( Figure 6A ) . The effect of Rho inhibition on translocation was also substantial ( Figure 6B ) . In various experiments , treatment with different batches of purified TAT-C3 ( 40μg/ml ) reduced YopE delivery into wild type-infected cells , by 40% to 75 % . Similar results were obtained when we tested the effect of C3 treatment on YopH translocation ( Figure 6B ) , indicating that the requirement of Rho for translocation is not a phenomenon restricted to YopE delivery . To test whether actin polymerization required for pore formation and translocation was dependent on Rho , we analyzed the effect of C3 on the induction of actin polymerization around the bacteria [18] . We found that the number of YopB-dependent actin halos was considerably reduced in the presence of C3 ( Figure 6C ) . To determine whether Rho is activated by infection with Y . pseudotuberculosis , we infected HeLa cells with strain yopEHJ ( YP27 ) for 5 , 10 , 15 and 20 min and we analyzed the amount of active Rho ( GTP-Rho ) in the cell lysates by a GTP-Rho pull-down assay , as described in Material and Methods . A peak of Rho activation was detected between 10 and 15 min after infection ( Figure 7A ) . A 15 min infection period was selected to test the levels of GTP-Rho induced by infection with yopEHJ ( YP27 ) , yopEHJB ( YP29 ) , yopEHJ , yadA , invD911E ( YP50/ invD911E ) , and yopEHJB , yadA , invD911E ( YP51/invD911E ) . Compared to YP27-infected cells , cells infected with YP29 have reduced amounts of GTP-Rho , indicating that Rho activation is YopB-dependent ( Figure 7B ) . Low affinity interaction with β1 integrin receptors by infection with YP50/ invD911E cause a reduced activation of Rho . However , YopB-independent Rho activation in YP29-infected cell lysates was greater than that of cells infected with YP51invD911E . This small difference , attributed to wild type invasin or YadA interacting with β1 integrin receptors , was consistent in three independent experiments . Overall , these experiments lead us to conclude that Y . pseudotuberculosis activates Rho by a process that involves YopB and high affinity interaction with β1 integrin receptors .
The TTSS-mediated translocation of bacterial effectors into host cells is an intricate mechanism that , although extensively studied , has not been completely unraveled [31] . Here we have found that Y . pseudotuberculosis engages the small GTPase Rho to control the delivery of effectors to the host cell . Activation of this signaling pathway is mediated by the YopB/YopD translocon in cooperation with the high affinity binding of invasin or YadA to β-1 integrins . It has been put forward that pore formation and translocation of effector Yops into the host cells are not related processes [19 , 32] . Pore formation has been recently implicated in mediating a caspase-1 dependent type of cell death in Salmonella-infected macrophages [21] . Shin and Cornelis [33] have recently reported that insertion of translocation pores in macrophages infected with a multi-effector mutant of Y . enterocolitica triggers activation of caspase-1 . Here we ruled out that in our infection system , YopB/YopD-mediated pore formation induces caspase-1 dependent cell death . Thus , amounts of a specific caspase-1 inhibitor large enough to block IL-1 β production in macrophages , does not prevent LDH release in Hela cells . Also , glycine treatment that efficiently prevented cell lysis in Salmonella infected macrophages failed to inhibit LDH release in Yersinia-infected HeLa cells . Based on these findings , we sustain that in our experimental system pore formation-induced LDH release is related to the process of Yop translocation . Both pore formation and translocation require activation of small Rho GTPases , as glucosylation of Rho , Rac and Cdc42 by C . difficile toxin ToxB potently inhibits the two processes . We found that Rac activation is not likely to be involved in pore formation or translocation . Thus over-expression of a dominant negative form of Rac does not prevent uptake of membrane impermeable dyes in cells infected with the pore forming strain . In line with these results , a specific Rac inhibitor , NSC23766 , that efficiently blocks Rac-mediated internalization , does not inhibit pore formation or translocation . On the other hand , we found that signaling downstream of Rho is essential for the control of Yops delivery . Treatment with C . botulinum C3 toxin , that converts endogenous Rho A , B and C into dominant negative forms [3] , potently down-regulates pore formation and translocation without affecting bacterial adhesion or internalization considerably . The type of host cell processes that Rho proteins regulate to promote translocation and pore formation most likely involves actin cytoskeleton rearrangements . Thus treatment with 2μg/ml actin polymerization inhibitor CD blocks pore formation [18] and decreases the level of YopE translocation by more than 60% . In early studies aim at demonstrating that Yop translocation is mediated by extracellular bacteria , Sory et al studied the effect of 5μg/ml CD treatment on the delivery of Yop-cyclase fusion proteins by Y . enterocolitica into murine macrophages [34] . Compared to the dramatic effect on bacterial uptake ( 2000 fold inhibition ) , the authors suggest that Yop translocation was not sensitive to the action of CD . However , their results show that CD treatment decreased YopE-cyclase translocation by 32% and YopH-cyclase by 52% . Using 10 times less CD ( 0 . 5μg/ml for 30 min ) and using a strain of Salmonella ectopically expressing YopE , Rosqvist et al reported that Yop translocation into HeLa cells was notably decreased [35] . The authors also reported that the same was observed when YopE was delivered by Y . pseudotuberculosis . Interestingly , our findings strongly suggest that actin polymerization required for pore formation and translocation is dependent on Rho , as inhibition of Rho A , B and/or C results in a decrease of the number of actin halos . Adhesion of bacteria to host cell is crucial for the activation of the TTSS . In Y . pseudotuberculosis two main adhesins , invasin and YadA , mediate tight binding to host cells by interaction with β1 integrin receptors . Here we show that in an inv/yadA mutant , constitutive expression of the pH6 antigen confers good adhesion properties to host cells . In spite of that , we found that such mutants are defective in pore formation and Yop translocation , suggesting that interaction with β1 integrin receptors is essential for the two processes . Mota et al . have shown that a minimal needle length is required for efficient functioning of the Yersinia injectisome , and that this length correlates with the length of the YadA protein [36] . We considered that the attachment imparted by pH6 antigen in the absence of invasin and YadA , might not provide that critical length . Our data suggest that this is not likely to be the case in our experimental system . First , a Y . pseudotuberculosis strain expressing pH6 antigen is able to stimulate a YopB-dependent proinflammatory response , including activation of NFκB and ERK , and production of IL-8 . Second , a single amino acid substitution in invasin ( invD911E ) , that is not expected to change its length , failed to mediate efficient Yop translocation . This mutant promotes adhesion without inducing receptor clustering and subsequent β1 integrin-mediated signal transduction . Altogether , these results suggest efficient translocation requires high affinity binding of β1 integrin receptors and subsequent activation of signaling . It is still conceivable that , independent of integrin signaling , tight bacterial adhesion mediated by high affinity interaction with β1 receptors preconditions effective translocation . The fact that interfering with β1 integrin signaling by the action of a Src inhibitor impairs efficient translocation , would argue against that idea . Still , we cannot discard the possibility that Src activity might also be required for YopB/D-dependent Rho activation . We predict that upon integrin clustering , RhoA could be recruited and generate a signal that polymerizes actin . It is well documented that invasin engagement of β1 integrin receptors triggers Rac1-mediated signals that induce bacterial internalization into epithelial cells [15] . This Rac1-mediated mechanism involves Arp2/3 , PIP 4 , 5 and capping-proteins [30] . Results from our GTP-Rho pull down assays suggest that bacteria producing invasin and YadA ( YP29 ) can also mediate Rho activation in a YopB-independent manner . There are further evidences in the literature that engagement of β1 integrin receptors can stimulate RhoA activation . Wong and Isberg have shown that RhoA is recruited at the nascent phagosome in Cos1 cells infected with a yopE yopT mutant of Y pseudotuberculosis [26] . Werner et al have reported that interaction of invasin-coated beads with α5β1 integrin in synovial fibroblast results in beads uptake by a process that is RhoA-dependent [37] . Also , activation of RhoA by engagement of α5β1 integrins by Ipa invasins has been implicated in the internalization of Shigella to HeLa cells [38 , 39] . Alternatively , β1 integrin may indirectly facilitate Rho activation by a focal adhesion kinase ( FAK ) -dependent pathway . Such a mechanism of Rho activation has been described for the regulation of microtubules stabilization at the leading edge of mouse fibroblasts [40] , and involves targeting of Rho to GM1-rich domains in the plasma membrane , where it can interact with downstream effectors . We envision a model in which high affinity binding to β1 integrin receptors , in addition to stimulating Rac activation , triggers Rho activation ( Figure 8 ) . Subsequently , YopB/D insertion into the plasma membrane stimulates increased Rho activation , and the cooperative activation of Rho stimulates Yop translocation . A central question is how Rho activation regulates Yop translocation . We hypothesize that Rho signaling induces changes in the host cell , such as actin polymerization , that are required for an efficient translocation process . One possibility is that , cell molecules present in specialized membrane microdomains , such as lipid rafts , are required for efficient translocation . These membrane microdomains would be recruited at the site of bacteria-host cell contact , as a result of Rho GTPases activation and actin polymerization . More injectisomes could then interact with lipid rafts at the site of bacteria , and more effector Yops would be translocated . Once proper amounts of Yops are delivered into the host cell , the process would be shut down to avoid further cell damage caused by excessive signaling . We based our hypothesis , in part , on the fact that Salmonella and Shigella-YopB homologues bind to cholesterol [41] , and that lipid raft are required for translocation in Salmonella , Shigella and EPEC [41] . Interestingly , actin polymerization and Rho GTPases activation have been shown to be involved in lipid raft clustering in B cells [42] , T cells [43] and NK cells [44] . Why is Rho-dependent , but not Rac-dependent , actin polymerization required for translocation ? Rho GTPases transmit signals that control the formation of distinct cytoskeletal structures through the interaction with different nucleating machineries . Cdc42 and Rac mediate nucleation of branched actin filaments through the Arp2/3 protein complex , leading to lamellipodia formation . On the other hand , Rho proteins stimulate unbranched actin filaments formation , such as those in stress fibers , via interaction with formins . It could be speculated that only F-actin structures generated by formins are important for translocation . The effect of the expression of dominant negative mutants of the formin mDia1 on translocation will be investigated in future studies . Findings from two studies that investigate translocation of TTSS effector proteins by Salmonella and Shigella in real time [45 , 46] indicate that effector translocation occurs right after host cell contact , with a half maximal rate of about 4 min . In our experimental model we detect the strongest Rho activation after 10 to 15 min infection with a YopEHJ bacteria . This is probably due to accumulation of GTP-Rho in the absence of the Rho inhibitors YopE and YopT . The decrease in the levels of GTP-Rho after 15 min is presumably due by the action of endogenous GAPs . We envision that during infection with wild type bacteria , the kinetics of Rho activation would be much faster . Translocation of Salmonella SipA and SopE , and Shigella IpaC were found to follow a linear kinetic [45 , 46] . Interestingly , however , slopes of IpaB secretion kinetics curves seemed to vary at different time points , suggesting that the speed of injection changes during the course of the translocation process resembling a slow-fast-slow type of mechanism . This type of translocation kinetic is what we would expect in our model . How does our model fit with the mechanism of Yop translocation in Y . pestis ? Although closely related to Y . pseudotuberculosis , Y . pestis lacks invasin and YadA . Unless Y . pestis has yet-unidentified adhesins that interact with β1 integrin receptors , we envision that the bacteria would activate Rho only by the stimulus elicited by YopB/D . In this situation , Rho activation would be limited , and therefore , one should expect that Y . pestis would be less effective for Yop translocation . A recent report suggests that , in macrophages , Y . pestis translocate less YopJ than a Y . enterocolitica strain expressing invasin and YadA [47] . However , in this report the authors suggest that this is most likely due to a difference between the YopJ protein from the two Yersinia species . We have preliminary results suggesting that Y . pestis deliver much less YopE in HeLa cells than Y . pseudotuberculosis . It has been proposed that , because bacterial effectors are directly injected within cell cytosol , the TTSS does not need to trigger signals through cell surface receptor [48] . Our data suggest that , although not essential , signal stimulated by engagement of β1 integrin receptors greatly enhances Yop translocation .
The wild-type serogroup III Y . pseudotuberculosis strain YP126 [49] , and the mutants derived thereof are shown in Table 1 . YP126 and its derivatives carry a naturally occurring deletion in virulence plasmid that inactivates the yopT gene and are thus devoid of YopT activity [50] . YP202/YP29 ( yopEHJB , inv ) was constructed by inserting the virulence plasmid of YP29 into a plasmid cured , inv::kan strain ( YP202 , Table 1 ) . To create YP50 ( yopEHJ , yadA , inv ) and the corresponding YopB-deficient mutant ( YP51 ) , the wild type yadA gene in YP202/pYP27 ( yopEHJ , inv ) and in YP202/pYP29 ( yopEHJB , inv ) , respectively , was replaced by yadA containing a frame shift deletion ( yadAfs ) , as follows . yadAfs was constructed by amplifying yadA with primer YadA F1 ( 5′-CCC GGG TTT GTA GTG GGC TGA CTC CGA C-3′ ) and B1 ( 5′'-GGC TGA ACT GGC TAA ACC TTT G-3′ ) . The yadA DNA fragment was subsequently blunt-cloned into pETBlue ( Novagen ) . QuikChange Site-Directed mutagenesis ( Stratagene ) was used to create the frame-shift and generate a SphI restriction site using primers F2 ( 5′-CA CAA GGT CCA GAA AAA AAA GAG CAT GCA TTA GCA GAA GCA ATA C-3′ ) , and B2 ( 5′-GTA TTG CTT CTG CTA ATG CAT GCT CTT TTT TTT CTG GAC CTT GTG-3′ ) . Plasmid pETBlue-yadAfs was digested with XmaI and subcloned into the suicide plasmid pSB890 containing sacB and TetR genes [51] . pSB890yadAfs was then introduced into S17-λpir and conjugated into CamR YP202/pYV27 and YP202/pYV29 . TetR CamR colonies were grown for several generations in the absence of Tet and were selected against the sacB on LB-5% sucrose . SucroseR , CamR and TetS colonies were screened for yadAfs by PCR using primers YadA F1 and B1 , followed by SphI-digestion of the amplified yadA fragment . A plasmid constitutively expressing pH6 antigen fimbriae , pAY66 ( LacP::psaABC , Table1 ) , a gift from R . Isberg ( Tufts University ) , was inserted into YP50 and YP51 by electroporation . To create YP54/pAY66 ( yopHJ , inv , yadA/psaABC ) , we replaced yopE::kan in YP50/pAY66 by wild type yopE , by allelic exchange using suicide plasmid pSB890YopE , essentially as described above . To construct YP50invD911E ( yopEHJ , yadA , invD911E ) and YP51invD911E ( yopEHJB , yadA , invD911E ) , the virulence plasmid from YP50 and YP51 were introduced into YPIII P− invD911E ( Table 1 , gift from R . Isberg ) by electroporation . To create YP54invD911E , we replaced yopE::kan in YP50invD911E by wild type yopE , by allelic exchange using pSB890yopE as described above . Plasmid pMMB67HEYadA ( pYadA ) [52] , was inserted into YP50 and YP54 by electroporation . HeLa cells were cultured as previously described [16] . For experiments carried out in the presence of inhibitors , cells were pre-incubated with 50–100 μM Ac-YVAD-cmk ( Calbiochem ) , 5mM Glycine ( Roche ) , 40ng/ml Clostridium difficile ToxB ( Calbiochem ) , 3 . 9 μM ( 2μg/ml ) cytochalasin D ( Sigma ) , 10 μM PP2 ( Sigma ) , 100 μM NSC23766 ( Calbiochem ) , 10 , 20 , or 40 μg/ml TAT-C3 . Bacteria used for infections were grown in Luria-Bertani ( LB ) broth either under conditions that stimulate ( low Ca2+ at 37 °C ) or repress ( high Ca2+ at 28 °C ) Yop expression [4 , 51] , at a multiplicity of infection of 50 to 100 . The plates containing the infected cells were centrifuged for 5 min at 700 rpm and incubated at 37 °C with 5% CO2 for different periods of time to allow bacterial-host cell interaction . Cells cultured in 24-well plates with coverslips were infected for 3 h with bacteria grown under low calcium conditions . A green fluorescent membrane-permeable nucleic acid stain ( SYTO10 ) and a red membrane-impermeable nucleic acid dye that label only cells with compromised membranes , ethidium homodimer-2 ( EthD-2 ) were provided in the DEAD-LIVE kit ( Invitrogen ) . After washing , a mixture of the two dyes was added to the wells and incubated in the darkness for 15 min at room temperature . Cells were then washed and fixed with 2% paraformaldheyde in PBS . Coverslips were mounted with 8 μl of ProLong mounting medium ( Molecular Probes ) and slides were then examined by immunofluorescence microscopy . Samples of culture media from wells containing infected cells were collected 3 h post infection . Levels of LDH were assayed using the CytoTox 96 assay kit ( Promega ) as previously described [16] . HeLa cells cultured in 6 cm2 dishes were infected with bacteria grown at high Ca2+ conditions . Infected cells were lysed with 0 . 2 ml of cold 1% Triton X-100 buffer as described [4] . Soluble and insoluble fractions were subjected to immunoblotting using an affinity purified polyclonal anti-YopE and anti YopH antibodies , as described previously [4] . Anti-β actin antibody was used as a loading control . Anti-rabbit antibodies conjugated with IR800 or IR680 were used as secondary antibodies , and the infrared signal was detected using an infrared imaging system ( Odyssey , LI-COR ) . Quantification of a fluorescent signal is more accurate than that generated by chemiluminescence because its intensity is not time-dependent . The bands intensities were calculated using the software provided by the Odyssey system , and the values were expressed as the YopE/β-actin ratio and plotted on a graph . Supernatants of infected HeLa cells were assayed for IL-8 production by ELISA ( Antigenix America ) five h after infection , as previously described [16] . Values obtained from triplicate wells were assayed in duplicate and averaged . HeLa cells were seeded onto glass coverslips at 105 cells per well in a 24-well tissue culture plate 24 h before infection . Cells were infected with bacteria at a calculated MOI of 50:1 . After a brief centrifugation step ( 5 min at 100 g ) , the plates were incubated for 30 min at 37 °C in a 5% CO2 incubator . A double-label immunofluorescence assay was used to differentiate between extracellular and intracellular cell-associated bacteria as previously described [4] . Coverslips containing infected cells were washed with PBS and fixed in 2% paraformaldehyde for 15 min . The washed coverslips were incubated with polyclonal anti-Yersinia antibody SB349 ( diluted 1:1000 ) for 40 min to stain extracellular bacteria . Washed coverslips were incubated for 40 min with FITC-conjugated goat anti-rabbit IgG diluted 1:250 . After washing , cells were permeabilized with 0 . 2% Triton X-100 for 10 min . Coverslips were washed and incubated with SB349 ( 1:1 , 000 ) for 40 min to label both extracellular and intracellular bacteria . Samples were then washed and incubated for 40 min with TRITC-conjugated goat anti-rabbit IgG ( 1:300 ) . All antibodies were diluted in PBS containing 3% BSA , and washes were conducted three times for 5 min with PBS containing 1% BSA . Coverslips were washed with PBS before mounting and examined by immunofluorescence microscopy . The percentage uptake was calculated as the number of [intracellular bacteria ( red ) /total bacteria ( green and red ) ] × 100 . The effect of Rac and Rho inhibitors on the formation of actin cups was tested in Hela cells seeded on coverslips . To inhibit Rac , the cells were treated for 6 h with NSC23766 ( 100μM ) in 5% serum-DMEM , or 5% serum-DMEM alone . To inhibit RhoA , B and C , TATC3 ( 40 μg/ml ) was added to the cells in serum free medium for 4 h , and control cells were incubated in serum free conditions for the same time . Hela cells were then infected for 10–15 min , washed and fixed as described above for the bacterial uptake assay . Double immunofluorescence was performed as detailed above for the bacterial uptake assay , with the addition of 50 U/ml of Rhodamine Phalloidin ( Molecular Probes ) together with the last secondary antibody . Images were captured with a confocal laser microscope . The percentage of bacteria ( extracellular and intracellular ) surrounded by an “actin halo” was calculated by counting a minimum of 150 bacteria . Plasmid pTAT–C3 ( a gift from Dafna Bar Sagi , Stony Brook University , NY ) was introduced into E . coli ( strain BL21 ) , and His-tagged-TAT–C3 protein were expressed by IPTG induction ( 1 mM IPTG , 4 h ) . Recombinant His-TAT–C3 was extracted from E . coli BL21 strain by sonication , and purified by fast protein liquid chromatography ( FPLC ) , as follows . The supernatant of the cell lysate was injected onto a Hi-trap Ni-column ( Pharmacia Co . ) . The column was washed with a 5 mM imidazole buffer solution and eluted using a gradient concentration of 1M imidazole buffer . After dialysis against PBS/0 . 5M NaCl , the purity of each TAT–C3 preparation was determined on polyacrylamide gels stained with Coomassie blue . Cells were seeded in 10 cm dishes at 90% confluency and were left uninfected or were infected at a moi:100 for different time periods . Cells were lysed in lysis buffer ( Upstate , Rho activation assay ) containing 10% glycerol , and protease inhibitor ( Roche ) . Cell lysates were clarified by centrifugation at 13 , 000 rpm at 4 °C for 10 min , and the supernatants were incubated with 30 μg of GST fused to the Rho binding domain of rhotekin bound to with glutathione beads , at 4 °C for 45 min . The beads were washed twice with lysis buffer and subjected to SDS-polyacrylamide gel electrophoresis on a 12% gel . Bound RhoA was detected by Western blot using a monoclonal antibody against RhoA ( Santa Cruz Biotechnology ) .
The GenBank ( http://www . ncbi . nlm . nih . gov/Genbank/index . html ) accession number for invasin is M17448 .
|
The type III secretion system ( TTSS ) is essential for the virulence of a number of Gram-negative human pathogens of enormous clinical significance . The molecular mechanisms by which TTSS effector proteins are translocated into the host cell are not well understood . The work presented here proposes a new model in which the enteropathogen Yersinia pseudotuberculosis manipulates the host cell machinery to control effector translocation . This involves activation of the host cell Rho GTPase by the cooperative action of adhesin-mediated high affinity binding to specific cell receptor molecules known as β1 integrins , and interaction of components of the TTSS with the host cell membrane . This molecular mechanism of controlling TTSS may not be restricted to Y . pseudotuberculosis and might take place during infection of host cells with other pathogens that encode homologues of Yersinia TTSS proteins . Our findings provide a good starting point to study the molecular nature of the complex interaction between bacterial pathogens bearing TTSSs and the host cell . Importantly , components that act by modulating the TTSS are potential targets for novel antimicrobials .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"microbiology",
"homo",
"(human)",
"eubacteria"
] |
2008
|
Yersinia Controls Type III Effector Delivery into Host Cells by Modulating Rho Activity
|
Hydrophobins , produced by filamentous fungi , are small amphipathic proteins whose biological functions rely on their unique surface-activity properties . Understanding the mechanistic details of the multimerization process is of primary importance to clarify the interfacial activity of hydrophobins . We used free energy calculations to study the role of a flexible β-hairpin in the multimerization process in hydrophobin II from Trichoderma reesei ( HFBI ) . We characterized how the displacement of this β-hairpin controls the stability of the monomers/dimers/tetramers in solution . The regulation of the oligomerization equilibrium of HFBI will necessarily affect its interfacial properties , fundamental for its biological function and for technological applications . Moreover , we propose possible routes for the multimerization process of HFBI in solution . This is the first case where a mechanism by which a flexible loop flanking a rigid patch controls the protein-protein binding equilibrium , already known for proteins with charged binding hot-spots , is described within a hydrophobic patch .
Hydrophobins are small ( 7–15 kDa ) proteins produced by filamentous fungi . They are globular and rigid proteins containing four disulfide bridges which stabilize the structure . Hydrophobins perform a variety of biological roles at interfaces that help fungi to adapt to their environment including adhesion and coatings of spores . Moreover , hydrophobins lower the surface tension of water so that fungal hyphae can penetrate the air-water interface and grow outside aqueous media [1–3] . The remarkable surface-activity properties of hydrophobins come from their amphiphilic nature . Besides their amphiphilicity , specific intermolecular interactions also contribute to their functional properties [4–9] . Due to their unique properties , hydrophobins have become attractive for use in several types of biotechnical applications . These include stabilization of colloidal dispersions , reverse the wettability of surfaces , dispersion of insoluble drug compounds , production of stable foams , and protein immobilization [8 , 10–13] . Hydrophobins are very soluble in water up to 100 mg/mL and display unusual detergent-like behaviour in solution as they form different kinds of oligomers , depending on the conditions and on the hydrophobin type [9 , 14 , 15] . Hydrophobins have been divided into two classes , class I and class II , based on the hydropathy profile of the amino-acid sequence [16] . In particular , class I hydrophobins are more resistant to dissociation using solvents and detergents than class II hydrophobins . Furthermore , class I hydrophobins tend to form rodlet-like aggregates at interfaces , whereas for class II hydrophobins various needle-like crystals and structured surface films have been observed [17–19] . The work described here was done on HFBI , a class II hydrophobins of the fungus Trichoderma reesei . The crystal structure of HFBI from T . reesei , solved in 2006 by Hakanpää and colleagues ( PDB id: 2FZ6 ) , shows four molecules in the asymmetric unit [20] . A tetrameric structure was also found in solution , where HFBI forms oligomers ( dimers and tetramers ) in a concentration-dependent manner . In solution , the tetramer is slightly larger and more elongated , with its monomers not as tightly packed as in the crystal . The oligomers are in some ways analogous to micelles , however , with the clear difference that the HFBI oligomers contain only two or four molecules [4] . Above a critical concentration ( 20 μM ) , HFBI is mainly in tetrameric form [9] . Besides oligomers , HFBI shows strong surface activity . HFBI is indeed a protein that self-organizes to form precise membrane structures [4 , 18 , 19 , 21 , 22] . Hydrophobin multimerization was suggested to protect the hydrophobic parts and that these associations disassemble at the interface to form monolayers . At the interface , HFBI exists as monomers , oligomers and surface monolayers , and the equilibrium is shifted towards surface assemblies [9 , 20] . Powers and colleagues [23] have shown that the mechanism of protein tetramerization via dimers is evolutionally favored over tetramerization via monomers and trimers . It is likely that the multimerization process of HFBI involves combination of monomers to dimers with the successive combination of dimers to tetramers [9 , 14] . In the HFBI structure , there are two types of molecules with respect to the conformation of the second β-hairpin motif ( residues 60 to 66 ) . Molecules A and C had this area in a similar “closed” conformation while molecules B and D both possessed an “open” conformation . The central β-barrel structure , with four disulfide bridges , remains unchanged [20] , see Fig 1 . In this paper , “closed” conformation of monomeric units A and C is named c , while “open” conformation of molecules B and D , is called o . It was suggested that movement in the β-hairpin area was most likely driven by the formation of the HFBI tetramer [20] . In a recent computational study , it was found that dimers and tetramers encounter complexes only form when monomers are in c conformation [24] . This supports the idea of an induced conformational transition upon encounter complex formation . In this work we explored the multimerization process of HFBI in solution . The fundamental role of the last β-hairpin in the oligomeric assembly is unveiled using all-atoms metadynamics simulations and a plausible oligomerization pathway is proposed .
All the computational models of HBFI here considered are based on the X-ray structure from Trichoderma reesei , solved at 2 . 1 Å resolution ( PDB id: 2FZ6 ) [20] . This structure is an hetero-tetramer with each unit consisting of 75 residues . The monomers are characterized by a different position of the second β-hairpin ( residues 60 to 66 ) with respect to the central β-barrel . In particular , chains B and D are in the so called conformation o , with the second β-hairpin exposed to the solvent , while chains A and C are in conformation c , with the second β-hairpin closed to the protein core . In the models , the starting units correspond to a specific chain in the crystal . Monomer ( c ) is chain A; monomer ( o ) is chain D; dimer ( cc ) is chain C + chain A ( superposed on chain D of crystal ) ; tetramer ( cccc ) is chain A + chain A ( superposed on B-C-D ) ; tetramer ( cocc ) is chain A + chain B + chain A ( superposed on C-D ) ; and tetramer ( coco ) is the crystal structure , see S2 Fig . The chain subjected to metadynamic bias is given in bold typeface ( see section “Well-Tempered Metadynamics ( MetaD ) ” ) . For each system ( monomer/dimer/tetramer ) , we followed the simulation protocol described hereinafter . The protein was put in a dodecahedric box of TIP3P water molecules ensuring a minimum distance to the box edges of 1 nm . The monomeric systems are neutral , while the dimer and tetramer have positive charge due to the presence of the Zn2+ ions at the interface between chains A/B and between chains C/D . The proper amount of Na+ and Cl− ions was added to reach a ionic concentration of 150 mM and ensure final neutral systems ( see S1 Table ) . A steepest-descent minimization was applied to relax the solvent molecules around the solute . The equilibration was performed in two steps: the system was at first thermalized up to 300 K coupling the protein and the solvent to a V-rescale thermostat [25] ( τt = 0 . 1 ) in the canonical ensemble ( NVT ) . Then , we switched to the NPT statistical ensemble , performing 100 ps of MD at 300 K , coupling the system with a Parrinello-Rahman barostat [26] ( τp = 2 ) . After this initial phase the system was ready for productive MD simulations . Production runs were carried out in the NPT ( p = 1 bar , T = 300 K ) statistical ensemble . All bonds were constrained with LINCS [27] , allowing to use a time step set of 2 fs . Periodic boundary conditions were applied to the systems in all directions . PME method [28] was used to evaluate long-range electrostatic interactions ( pme order = 4 , fourier spacing = 0 . 12 ) , and a cutoff of 10 Å was used to account for the van der Waals interactions . Coordinates of the systems were collected every 2 ps . All MD simulations were carried out with GROMACS-5 [29] using the AMBER99 force field [30] on GPU/CPU machines . The length of the MD simulations was of 150 ns , for monomer ( c ) and monomer ( o ) , 100 ns , for dimer ( cc ) , and 300 ns , for tetramer ( cccc ) and tetramer ( coco ) ( see S1 Table ) . Within each monomeric unit four covalent crosslinks between CYS18-CYS48 , CYS19-CYS31 , CYS8-CYS57 , and CYS58-CYS69 ( see S1 Fig ) have been defined and treated according to the disulfide bridge parameterization as in AMBER99 force field [30] . Standard MD were used for guessing CVs for metadynamics simulations [31] and for all analysis other then the free energy calculations . Well-tempered metadynamics , labelled as MetaD , simulations were performed with GROMACS-5 [29] using AMBER99 force field [30] and the PLUMED version 2 . 2 [32] plugin for free energy calculations . The starting structure for MetaD were taken after the NPT equilibration described above . The collective variables ( CVs ) used to describe the transition between monomer ( c ) and monomer ( o ) were the distance δ = [ ASP CA 30 − GLN CA 65 ] , and the torsion τ = [ VAL C 59 − ALA N 60 − ALA C A 60 − ALA C 60 ] . Both variables were necessary to properly describe the transition c/o without irreversibly distorting the structure of the β-hairpin . Metadynamics bias was constructed adding a Gaussian function with an initial height of 1 . 2*T/T0 kJ/mol and a width of 0 . 1 . T0 was set to 300 K and the bias factor ( γ = ( T + ΔT ) /T ) was set to 10 . An upper wall at 1 . 3 nm with a κ of 2000 kJ/mol/nm2 was associated to the CV δ . This choice was justified by the fact that in the open conformation o , the value of δ is 1 . 2 nm . In multimers the metadynamic bias was applied only to chain D , highlighted in bold typeface when specified in the text . For example , performing MetaD on tetramer ( cccc ) means that the starting structure was a tetramer composed of four c conformations and the MetaD bias was applied to chain D . Convergence was checked by computing the free energy as a function of simulation time ( 10 ns blocks ) . Moreover , the value of FEP ( δ ) at δ min 3 = 1 . 25 nm nm has been plotted as function of time . At convergence , the reconstructed profiles should be similar , and the value at δ min 3 = 1 . 25 nm should be constant ( see S3 Fig ) . Each MetaD simulation is 200 ns long , enough to ensure a proper convergence of the free-energy ( all details in the SI ) . In order to obtain reference regions on the CVs space sampled by the c and the o forms , the joint probability density function f ( δ , τ ) has been computed from standard MD simulations of monomer ( c ) and tetramer ( coco ) . In the tetramer case , f ( δ , τ ) has been computed as an average across the two monomeric units in o form . Contour levels specifying the c and o regions on the FES plots ( black lines in Fig 2 ) are specified as volume percentages . For example , a contour at 90% encloses the 90% of the most probable data points and excludes the remaining 10% . The contour volume percentage can be specified as follow: given a joint probability density function fi ( δ , τ ) we want to find the set A which includes all points i such that ∑ i ∈ A f i ( δ , τ ) d δ d τ = χ where χ is , for example , 0 . 9 . In order to compute the contour volume percentage the following algorithm is used: i ) sort all points i according to the value of fi ( δ , τ ) in decreasing order obtaining the ordered list L = { i k } k = 1 k = N . ii ) Compute the cumulative sum on the sorted values , C = cumsum ( L ) . iii ) Compute Z = ∑i fi ( δ , τ ) . iv ) Set A is defined by all ik such that C ≤ 0 . 9Z . The isocontour line is defined by all ik such that C ≡ 0 . 9Z . Two-dimensional free energy surfaces as a function of δ and τ have been obtained by summation of the added Gaussian hills . The 2D surface was discretized using a spacing of 0 . 035 nm and 0 . 035 deg on δ and τ , respectively . Free energy profiles as a function of a single CV , FEP ( δ ) and FEP ( τ ) , have been computed integrating out one CV from the two-dimensional FES ( δ , τ ) ( Fig 3A and 3B ) . The FEP ( δ ) as a function of simulation time ( every 10 ns blocks ) was computed and the last 5 blocks were used to estimate the mean 〈 F E P ( δ ) 〉 = 1 N ∑ i 5 F E P i ( δ ) and the standard error of the mean as s e F E P ( δ ) = σ F E P ( δ ) n , where σFEP ( δ ) is the standard deviation across the five simulation blocks ( n = 5 ) . Throughout the paper , the angular brackets for the average FEP were dropped for clarity . A similar procedure was applied for the other CV , τ . On FEP ( δ ) , three free energy minima have been selected as representative of c conformation ( δ min 1 = 0 . 47 nm ) and o conformation ( δ min 2 = 0 . 81 nm and δ min 3 = 1 . 25 nm ) . The free energy values at the three minima have been computed for monomer ( c ) , dimer ( cc ) , tetramer ( cccc ) , and tetramer ( cocc ) and plotted as mean ± the 95% confidence interval , C I 95 % = Δ G ( δ min i ) ± 1 . 96 s e . Hydrogen bonds were calculated using GROMACS-5 software tools on the 300 ns standard MD simulations for tetramer ( cccc ) and tetramer ( coco ) . The H-bond persistence was computed as the number of times the ith H-bond was found , divided by the total number of frames . Only H-bonds with persistence > 5% were retained for the analysis . We decomposed the hydrogen bonds into four groups: i ) intra-hairpin , ii ) intra-chain , iii ) inter-chain , and iv ) hairpin-solvent . The intra-hairpin includes hydrogen bonds formed within the residues 60–66 of the β-hairpin . The intra-chain group corresponds to the hydrogen-bonds between the β-hairpin and the rest of the chain . The inter-chain group contains hydrogen-bonds established between the β-hairpin and the chain facing the β-hairpin . Finally , hairpin-solvent group includes hydrogen bonds formed by the residues of the β-hairpin and the solvent ( see Fig 4D for a description of the groups ) . While for groups i , ii and iii an atomistic detail was considered , for group iv , the average number of hydrogen bonds formed between a given aminoacid and the water was used . H-bonds analysis was performed on one monomeric unit ( chain D ) within tetramer ( cccc ) as well as tetramer ( coco ) . Solvation free energy has been computed using software gmmpbsa [33] . Briefly , the solvation free energy is expressed as sum of two terms Gsolvation = Gpolar + Gnon − polar . Gpolar is obtained solving the linearized Poisson-Boltzmann equation using APBS software [34] . Ionic strength was set to 150 mM , solute and solvent static dielectric constants were set to 2 . 0 and 78 . 4 respectively . Gnon − polar was computed using the solvent accessible surface area ( SASA ) model [33] as Gnon − polar = γSASA + b , where γ is a coefficient related to surface tension of the solvent and was set to 0 . 0226778 kJ mol−1 Å−2 , and b = 3 . 84982 kJ/mol is a fitting parameter . Hundred equally spaced frames were extracted from standard NPT molecular dynamics simulations of monomer ( c ) , monomer ( o ) , dimer ( cc ) , dimer ( co ) , tetramer ( cccc ) , and tetramer ( coco ) . Frames were separated by at least 1 ns ( depending on total simulation length , see S1 Table for the simulations details ) from each other in order to avoid correlations . ΔGpolar and ΔGnon − polar terms were computed on each frame . Statistical analysis was performed comparing pairs monomer ( c ) /monomer ( o ) , dimer ( cc ) /dimer ( co ) , and tetramer ( cccc ) /tetramer ( coco ) using a Welch’s t-test . p<0 . 01 was considered statistically significant . The electrostatic potential was obtained solving the linearized Poisson-Boltzmann equation using APBS software [34] . Ionic strength was set to 150 mM , solute and solvent static dielectric constants were set to 2 . 0 and 78 . 4 respectively . The single sphere Debye-Hükel model was used as boundary condition for coarse grid . Smoothed molecular surface was used to define the dielectric boundaries . The electrostatic potential has been computed separately for chains A-B-C and chain D in tetramer ( cccc ) and tetramer ( cocc ) , chain C and chain D , separately , in dimer ( cc ) . A cluster analysis was performed on standard MD simulations ( see S1 Table for details about simulations parameters ) using single linkage algorithm setting 0 . 15 nm as RMSD cutoff . The centroid of the most populated cluster was used as reference structure for the calculation of the electrostatic potential maps in Fig 5A , 5B and 5C . In order to estimate a local electrostatic potential at the interface between chains C and D , the potential was averaged within a cuboid subregion enclosing the C/D interface . The subregion was defined as normal to the plane formed by the β-sheet of chain D at the C/D interface and with sides of length 2 . 0 , 2 . 0 , and 1 . 0 nm , see Fig 5A , 5B and 5C . This local electrostatic potential was computed over the entire standard MD trajectory using conformations every 1 ns . Mean value and standard error of the mean ( s . e . m ) have been then obtained , see Fig 5D . Interfaces between all monomeric units of tetramer ( cccc ) , tetramer ( cocc ) , and tetramer ( coco ) , have been computed from the entire standard MD simulations . The following chain pairs have been considered: A/B , B/C , C/D , A/D , B/D , and A/C . Each interface has been described in terms of interface area , distance maps , and residues at the interface . For the pairs of chains i/j the interface area has been computed as SASAi/j = ( SASAi + SASAj ) − SASAi , j , where SASAi , j is the SASA computed for the complex i/j , while SASAi and SASAj are the SASA of the isolated chains . Solvent accessible surface area was computed using the GROMACS tool sasa [29] . Distance maps have been obtained by measuring the smallest distances between residue pairs ( heavy atoms only ) for all trajectory frames and averaging over time . Interacting residues have been defined as pairs of aminoacids whose distance was up to 0 . 45 nm on the distance map [35] . The GROMACS tool mdmat [29] was used for this purpose . All data and statistical analysis were performed using the software package R version 3 . 2 [36] . Figures for the three-dimensional protein structures have been obtained using VMD version 1 . 9 . 2 [37] and Chimera version 1 . 10 [38] .
The role of the last β-hairpin in the oligomeric assembly was probed by exploring the transition from conformation c to o using metadynamics ( MetaD ) [31] . The MetaD bias was applied to two configurational collective variables ( CVs ) : the distance δ = [ ASP CA 30 − GLN CA 65 ] , and the torsion τ = [ VAL C 59 − ALA N 60 − ALA C A 60 − ALA C 60 ] ( Fig 1 ) . These CVs where empirically selected observing the β-hairpin motion , in standard MD simulations , of the monomer in open and closed forms . While the distance δ is clearly an obvious coordinate for describing the opening of the β-hairpin , torsion τ has been selected as this dihedral angle changes from ≈-150 deg to ≈-60 deg from the closed to the open conformation ( Fig 1 ) . As mentioned in method section , upper/lower bounds were added to these CVs to avoid the unfolding of the protein structure . In order to understand the influence of the multimerization process on the conformational rearrangement , four MetaD simulations have been performed starting from monomer ( c ) , dimer ( cc ) , tetramer ( cccc ) , and tetramer ( cocc ) conformations . Convergence of the MetaD simulations has been assessed as described in the Method Section . The free energy surfaces as a function of δ and τ , FES ( δ , τ ) , are reported in Fig 2 ( see also the corresponding probability density functions in S4 Fig ) . FES have been shifted as min ( FES ( δ , τ ) ) = 0 . At 300 K , the β-hairpin is flexible so , standard MD simulations have been performed for the monomer in conformation c ( monomer ( c ) ) and for the tetrameric crystal structure ( tetramer ( coco ) ) to obtain reference regions on the CVs space sampled by the c and the o forms . From these MD simulations , percentage volume contours enclosing 90% of the most probable conformations were plotted over the FES ( δ , τ ) in order to locate the c ( continuous line ) and o ( dashed line ) conformation . Considering the MetaD simulations of the monomer , a main minimum was found at δ min 1 = 0 . 47 nm , τ = [-150 , -60] deg which corresponds to the c form ( Fig 2A ) . In solution the equilibrium distribution of the HFBI monomer is shifted to the c conformation . The dimer shows a different behaviour , three main minima appears on the surface , at δ min 1 = 0 . 47 nm , τ = -150 deg; δ min 2 = 0 . 81 nm , τ = -150 deg; and δ min 3 = 1 . 25 nm , τ = -60 deg . A video showing the MetaD simulation of the dimer can be found in SI , S1 Video . The FES ( δ , τ ) of the homo-tetramer cccc is similar to the FES of the monomer in solution , i . e . the thermodynamically favoured state is the c conformation . MetaD simulations were performed starting from the hetero-tetramer cocc in order to assess for a cooperative effect in the conformational rearrangement ( c to o state ) of one monomeric unit depending on the presence of a second monomer in the o form . The FES ( δ , τ ) of hetero-tetramer resembles the one of the dimer , where multiple main minima exist . In particular , two broad minima are visible around δ min 1 = 0 . 47 nm , τ = -150 deg , and δ min 2 = 0 . 81 nm , τ = -60 deg . To summarize the differences between the four MetaD cases , free energy profiles as a function of a single CV , FEP ( δ ) , FEP ( τ ) have been computed integrating out one CV from the two-dimensional FES ( δ , τ ) ( Fig 3A and 3B ) . From there , it is clear the different behaviour of the monomer ( c ) and the tetramer ( cccc ) compared to the dimer ( cc ) or the tetramer ( cocc ) forms . To quantify those differences and assess their statistical significance , the values of the main free energy minima on the most representative collective variable , the distance δ , have been compared , Fig 3C . Considering the distance as unique CV , the c conformation is identified by δ min 1 = 0 . 47 nm , while the o conformation is defined by δ min 2 = 0 . 81 nm and δ min 3 = 1 . 25 nm . Monomer ( c ) and tetramer ( cccc ) have a pronounced minimum at δ min 1 = 0 . 47 nm while the other two distances ( δ min 2 = 0 . 81 nm , δ min 3 = 1 . 25 nm ) have large free energy values . Conversely , in the dimer ( cc ) the equilibrium distance is shifted toward δ min 2 = 0 . 81 nm and δ min 3 = 1 . 25 nm , i . e . the o form . In tetramer ( cocc ) , the profile is flatter with nearly zero free energy value for δ min 1 = 0 . 47 nm and δ min 2 = 0 . 81 nm which confirm an intermediate behaviour between dimer ( cc ) and tetramer ( cccc ) . In MetaD simulations , the monomeric units not subjected to MetaD bias remain in their initial configurational state ( c or o ) . This has been checked by plotting the values of δ and τ for chains A , B and C in the tetramers and chain C in the dimer for all MetaD simulations , see S5 Fig . Hydrogen bonds ( H-bonds ) formed by the aminoacids in the β-hairpin and the rest of the molecule affect the stability of the c/o conformations and can explain why the β-hairpin opens within dimer ( cc ) or tetramer ( cocc ) and remains closed in monomer ( c ) or tetramer ( cccc ) . We decomposed the H-bonds into four groups: i ) intra-hairpin , ii ) intra-chain and iii ) inter-chain , and iv ) hairpin-solvent , see Methods Section for details . In Fig 4A and 4B the H-bonds and their average persistence is shown in a chord diagram for group i , ii and iii , see also S2 Table for quantitative information about the H-bonds networks . Considering tetramer ( cccc ) , several persistent H-bonds are present between the β-hairpin and the rest of the chain , which is expected as the β-hairpin is parallel to a β-strand . Almost no H-bond is found within the β-hairpin itself . Two slightly persistent H-bonds form between the β-hairpin and the facing chain ( Fig 4A ) . Focusing on tetramer ( coco ) , it is clear that the drastic reduction of intra-chain H-bonds is due to the β-hairpin opening . Despite the persistence is low , several H-bonds form within the β-hairpin itself and some with the interfacing chain ( Fig 4B ) . In tetramer ( coco ) , the H-bonds persistence is lower and the average number of H-bonds is also smaller compared to tetramer ( cccc ) . However , looking at the average number of H-bonds formed by the residues within the hairpin and the solvent , the picture is inverted ( Fig 4C ) . The β-hairpin opening exposes its mainchain to the solvent allowing the formation of stable H-bonds with water molecules . In particular , approximately two H-bonds are gained for ALA60 , VAL62 and GLY64 in the transition from c to o conformation . Recalling that the stable conformation of the monomer in solution is the c form , the opening of the β-hairpin in dimer ( cc ) can not only depend on solvent mediated H-interactions , i . e . a large hydrophobic patch is present on the surface of the HFBI monomer and may affect its stability and the β-hairpin rearrangement . Polar ( ΔGpolar ) and non-polar ( ΔGnon−polar ) contribution to the solvation free energy have been calculated using APBS [34] on 100 structures extracted from equilibrium simulations . Non-polar contribution has been computed using the solvent accessible surface area model [33] ( further details in the Method Section ) . To assess for differences between c and o forms in different oligomerization states , ΔGnon−polar was calculated for the following pairs: monomer ( c ) /monomer ( o ) , dimer ( cc ) /dimer ( co ) , and tetramer ( cccc ) /tetramer ( coco ) ( Fig 6 ) . Pair monomer ( c ) /monomer ( o ) shows statistically significant ( Welch’s t-test , t = -8 . 9 , df = 196 . 5 , p<0 . 001 ) difference in ΔGnon−polar , with the o conformation having a large free energy value . Close to the β-hairpin , e . g . at the interface between chain C and chain D , the electrostatic potential varies depending on whether conformation of chain B is c or o and if chain B is present or not , see Fig 5 and S2 Video . In the very same region , the electrostatic potential of chain D in c form is also negative , creating an electrostatic clash between chains D and chain C . The local electrostatic potential at the C/D interface is negative in c− , and positive in coc− and ccc− , see Fig 5D . The presence of an electrostatic clash in the dimer may promote the loop opening , while the complementary electrostatic cloud in tetramer ( cccc ) keeps the loop closed . Tetramer ( cocc ) has an intermediate behavior having a positive local electrostatic potential similarly to tetramer ( cccc ) but lower in magnitude . The importance of long range interactions for the cooperative effect of the loop opening observed in tetramer ( cocc ) is also supported by the analysis of intra-chains interfaces ( see S7 Fig and S3 Table ) . In the tetramer , there are six possible contact interfaces between the four monomeric units . The interfaces between chains A/B and C/D maintained the same area while changing the contact residues , reflecting the β-hairpin rearrangement . The interfaces between chains A/D and B/C kept a constant area and the same contact residues . These interfaces were rather rigid , hence , they are not responsible for the cooperative transition . On the other hand , the contact areas between chains A/C and B/D shrank during the transition from tetramer ( cccc ) to tetramer ( coco ) or tetramer ( cocc ) . In particular , the interface between B and D , which was already small , disappeared , while interface A/C varied part of its contact residues . The variation of interfaces A/C and B/D depends upon a rigid rotation of the B-C chains with respect to the A-C chains and is not due to local rearrangements , see tetramer ( cccc ) in Fig 7 . As a consequence of this rotation , the electrostatic potential at the C/D interface couples with the β-hairpin rearrangement , as previously described . Moreover , in tetramer ( cccc ) the opening of the β-hairpin ( chain D or B ) may be hindered by steric effects due to the position of residues 20–29 ( chain C or A ) , see Fig 7 .
The goal of this study was to clarify the multimerization mechanism of HFBI in solution . HFBI forms dimers and tetramers in a concentration dependent manner . Above a critical concentration ( 150 g/L ) HFBI is mainly tetrameric [14 , 24] . The crystal structure of HFBI is also a tetramer which contains two types of molecules named in this work c and o conformations differing only by the position of the last β-hairpin motif . The rest of the molecule is exceptionally rigid , due to the presence of four disulfide bridges which stabilize the structure , and is almost identical among the four chains . Using Brownian dynamics simulations , it was found that dimers or tetramers encounter complexes only assemble from c conformations [39] . This finding supports the suggestion that the conformational rearrangement of the last β-hairpin found in the HFBI crystal structure is induced by tetramer formation [14] . The role of the last β-hairpin in the multimerization mechanism was assessed in this work by exploring the transition from conformation c to o in the monomer , dimer and tetramer using metadynamics . In dimers and tetramers the metadynamic bias was only applied to one monomeric unit , chain D ( see Methods for details ) . Throughout the manuscript , whenever monomer , dimers or tetramers are specified , the conformation of the monomeric units is given in parenthesis and the chain subjected to MetaD is given in bold typeface . At first , we investigated the preferred conformation of the HFBI monomer in solution . The FES of the monomer obtained from MetaD simulations shows a clear minimum in correspondence of the c form ( see Figs 2A and 3 ) . In solution , c form is thermodynamically favoured . Upon dimerization , multiple minima ( mainly three ) appear distinctly changing the FES . The minimum in correspondence to the o conformation , Figs 2B and 3 , is particularly relevant . These results indicate that , in the dimer , the c to o transition is allowed . In the c conformation , the last β-hairpin is involved in an anti-parallel β-sheet . Several H-bonds must be broken in order to move the β-hairpin to the o conformation . A possible explanation for the allowed transition within the dimer is the formation of a H-bond network which compensate for the loss of the H-bonds between the last β-hairpin and the β-sheet . In order to check for this , the H-bond network involving the β-hairpin in tetramer ( cccc ) and tetramer ( coco ) has been compared . In c conformation , 4 . 5 hydrogen bonds are present , on average , between the β-hairpin and the β-sheet ( Fig 4A , intra-chain group ) . In o conformation , only transient H-bonds are formed: persistence ≈10% and average number of H-bonds per frame less than 1 in all groups ( Fig 4B ) . The loss of H-bonds is not restored by new H-bonds within the protein . However , looking for H-bonds formed with the solvent , the exposed conformation of β-hairpin in o form allows several ( approximately 6 ) H-bonds to be established with water molecules , see Fig 4C . If the opening of the β-hairpin was due to solvent mediated H-bonds , the monomer in solution could have also been stable in o conformation , however this is not observed . The reason for the stabilization of the o form in the dimer does not only depend on the H-bonds network . In details , the HFBI monomer has a large non-polar patch exposed to the solvent . The opening of the β-hairpin may further increase the non-polar solvent exposed area , thus , destabilizing the molecule . This has been indeed proved by computing the non-polar contribution to the solvation free energy ΔGnon−polar for monomer ( c and o ) , dimer ( cc and co ) and tetramers ( cccc and coco ) . In the monomer the transition from c to o significantly increases the ΔGnon−polar due to the exposure of non-polar residues . In dimers or tetramers , part of the non-polar surface patch is buried by the presence of other monomeric units canceling out the differences between the homo/hetero-dimer and the homo/hetero-tetramer , see Fig 6 . This explains why the transition c/o is allowed in the dimer and not when HFBI is in the monomeric form . Tetramer formed by four c conformations should behave similarly to the dimer , however , the FES of tetramer ( cccc ) resembles the monomeric one , see Figs 2C and 3 . That is , the equilibrium conformation of the molecule within a homo-tetramer is the c form . In tetramer ( cccc ) the β-hairpin does not undergo a conformational rearrangement due to electrostatic and steric effects . At equilibrium , chains B-C in tetramer ( cccc ) rigidly rotate with respect to chains A-D , compared to the tetramer ( coco ) conformation ( Fig 7 ) . This rotation leads to a variation of the electrostatic potential at the C/D interface ( Fig 5 ) and to the formation of contacts that reduce the possibility of β-hairpin opening ( Fig 7 and S3 Table ) . This coupling between quaternary and tertiary structure rearrangements has been well studied , for example , in hemoglobin [40 , 41] . On the other hand , looking at the tetramer ( cocc ) , where one chain is already in conformation o , an intermediate behaviour between the monomer and the dimer can be observed in term of FES ( Figs 2D and 3 ) . In tetramer ( cocc ) , chain C keeps the same internal structure and the same relative orientation with respect to chain D as in the dimer ( S6 Fig and Fig 7 ) . However , the electrostatic potential at the C/D interface changes due to the presence of chains A and B as clear from Fig 5 and S2 Video . Changes in the electrostatic potential at the C/D interface are responsible for the lower probability of β-hairpin opening in tetramer ( cocc ) . In chain D , the region of the β-hairpin has a large negative patch , extending from the molecule surface . On the facing chain ( chain C ) a region with a negative electrostatic potential is also present , however the magnitude of this negative area changes depending on the presence/absence of chain B and on its conformation . In particular , in the dimer , where chain B is not present , the electrostatic potential has the largest magnitude . In tetramers , when chain B is in c conformation , the electrostatic potential at the interface between chain C and the β-hairpin is reduced . When chain B is in o conformation , the magnitude of negative electrostatic cloud is in between the dimeric and the homo-tetrameric one . The stronger repulsion exists in the dimer where the overlap of the two same-charged regions may promote the opening of the loop . On tetramer ( cccc ) , the repulsion is notably lower preserving the closed form while in tetramer ( cocc ) an intermediate behavior occurs , where the opening may happen however with low probability . Summarizing these findings , together with the knowledge that dimers and tetramers are present in solution [4 , 9] , the multimerization mechanism can be dissected , see Fig 8 . At first , unfavored routes are excluded . In particular , transition 9 and 10 ( see Fig 8 ) can not occur according to what found by Brownian dynamics simulations [24] , and because the monomer ( o ) is not stable in solution as found by MetaD simulations . We do not have enough information to determine the preferred direction of transition 7 and 8 . From Brownian dynamics simulations , tetramers in c forms have been observed , however , it is highly likely that they are transient encounter complexes . This idea is supported by the unfavored transition from c to o within a tetramer ( cccc ) found in this work . Two possible multimerization mechanism can now be proposed . The first one , the most probable route , implies the association of two monomers in c conformation into a dimer cc , transition 2 in Fig 8 . This association is supported by Brownian dynamics simulations results [24] where dimeric cc encounter complexes were found to be favoured over co and almost no dimers in oo conformation were found . Within the dimer , the conformational change of one molecule to o ( transitions 3 ) is largely favoured according to our findings . Then , it is possible that two co dimers can now assembly into stable tetramer ( coco ) ( transition 4 ) . No direct evidence is available for this last step , however , this route is consistent with the general finding that passing through dimers is evolutionary preferred [9 , 23] . Another possible route , is the association of one co dimer with one cc dimer into a tetramer ( cocc ) ( transition 5 ) . Then , the motion of the last chain to o conformation can occur according to the results of MetaD simulations ( transition 6 ) . This second mechanism is less probable compared to the first one as the cocc → coco transition is not as favoured as in the dimer . The small free energy differences in dimer ( co ) and tetramer ( coco ) imply that the β-hairpin can relatively easily go back and forth between the c and the o conformation . This can be also seen looking at the densities of the conformational states of tetramer ( coco ) in standard MD simulations . Already in 300 ns , the β-hairpin performs large movements passing by δ min 3 = 1 . 25 nm , δ min 2 = 0 . 81 nm , and δ min 1 = 0 . 47 nm on the distance coordinate . A complete transition from o to c , which implies the rotation of τ , is not however observed in standard MD . These findings allow to draw a biological role for the proposed association mechanism . As previously suggested [20] , hydrophobin multimerization is an efficient way to protect the large hydrophobic patch , i . e . avoid unwanted strong unspecific interactions . Nevertheless , in order to exploit their biological function ( e . g . lowering the water surface tension while the hyphae are growing [16] ) , multimers must not be overly stable: they have to dissociate at the air/water interface [9 , 20] . The motion of the last β-hairpin is essential to fine tune the stability of the HFBI multimers . It is highly likely that the arrangement of HFBI at the interfaces is also affected , as the hydrophobic interaction surface and lateral interactions are modified by the movement of the last β-hairpin . This result is remarkably important in order to clarify the mechanism of arranging at the interface and enhancing hydrophobin-based technological applications [42] . More generally , the strategy where a rigid patch flanked by a flexible region allows to adjust protein-protein interaction energy , was already found in other protein complexes [43] . However , the interface was composed of charged residues [43 , 44] . To the best of our knowledge this is the first example where this unique fine-tuning association mechanism occurs within a hydrophobic interface .
|
Fungi proliferate by creating a complex hyphal network growing within a wet environment . However , for most fungi to colonize new territories , they must produce spores carried by aerial hyphae and spread them into the air . Aerial structures need to overcome the surface tension of the surrounding water in order to grow into the air . This process requires hydrophobins , a remarkable class of self-associating fungal proteins which lower the surface tension at the air/water interface by creating a thin amphipathic layer . In solution they form multimers in equilibrium with the interfacial layer . Due to their unique surface-activity properties , hydrophobins have been used for a variety of biotechnical applications . We used enhanced sampling molecular dynamics simulations methods to study the multimerization process in solution of a hydrophobin from Trichoderma reesei ( HFBI ) . We clarified the fundamental role of a small flexible region within the HFBI monomer involved in the formation of multimers . A flexible loop flanking a rigid interaction patch is able to fine-tune the interaction energy . This mechanism , already known for charged binding patches , is described here for hydrophobic hot-spots . This result is remarkably important in order to clarify the mechanism of arranging at the interface and enhancing hydrophobin-based technological applications .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"dimers",
"(chemical",
"physics)",
"chemical",
"bonding",
"crystal",
"structure",
"electricity",
"condensed",
"matter",
"physics",
"fungal",
"structure",
"electrostatics",
"crystallography",
"thermodynamics",
"hydrogen",
"bonding",
"physical",
"chemistry",
"mycology",
"solid",
"state",
"physics",
"chemistry",
"biophysics",
"free",
"energy",
"physics",
"biochemistry",
"biochemical",
"simulations",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology",
"chemical",
"physics",
"biophysical",
"simulations"
] |
2016
|
Induced Fit in Protein Multimerization: The HFBI Case
|
Two-component signal transduction pathways comprising histidine protein kinases ( HPKs ) and their response regulators ( RRs ) are widely used to control bacterial responses to environmental challenges . Some bacteria have over 150 different two-component pathways , and the specificity of the phosphotransfer reactions within these systems is tightly controlled to prevent unwanted crosstalk . One of the best understood two-component signalling pathways is the chemotaxis pathway . Here , we present the 1 . 40 Å crystal structure of the histidine-containing phosphotransfer domain of the chemotaxis HPK , CheA3 , in complex with its cognate RR , CheY6 . A methionine finger on CheY6 that nestles in a hydrophobic pocket in CheA3 was shown to be important for the interaction and was found to only occur in the cognate RRs of CheA3 , CheY6 , and CheB2 . Site-directed mutagenesis of this methionine in combination with two adjacent residues abolished binding , as shown by surface plasmon resonance studies , and phosphotransfer from CheA3-P to CheY6 . Introduction of this methionine and an adjacent alanine residue into a range of noncognate CheYs , dramatically changed their specificity , allowing protein interaction and rapid phosphotransfer from CheA3-P . The structure presented here has allowed us to identify specificity determinants for the CheA–CheY interaction and subsequently to successfully reengineer phosphotransfer signalling . In summary , our results provide valuable insight into how cells mediate specificity in one of the most abundant signalling pathways in biology , two-component signal transduction .
Bacteria , Archaea , and some eukaryotes use two-component signalling pathways to detect environmental conditions and bring about appropriate changes in cellular behaviour [1] , [2] . Two-component pathways comprise sensor histidine kinases ( HPK ) and response regulators ( RRs ) . Environmental stimuli control the rate at which the HPK autophosphorylates on a conserved histidine residue . Once phosphorylated , the HPK transfers the phosphoryl group to an aspartate residue within the receiver domain of the cognate RR . The phosphorylated RR ( RR-P ) , often a transcriptional regulator , then effects a response appropriate to the original stimulus . Some bacteria have over 150 different HPK and RR pairs , and the specificity of the phosphorylation reactions between them needs to be tightly controlled to prevent HPKs from inappropriately phosphorylating and activating noncognate RRs . A number of mechanisms contribute to this specificity [3] , [4] , although the primary one is molecular recognition , in which a HPK shows a strong kinetic preference for its cognate RRs [5] , [6] . Understanding the mechanisms involved in molecular recognition will not only allow prediction of interacting pairs , but potentially allow the rewiring of bacterial sensory pathways for use in synthetic biology . Many two-component systems utilize an additional element , the Hpt ( histidine-containing phosphotransfer ) domain in phosphotransfer; these include multistep phosphorelays and chemotaxis signalling pathways . In multistep phosphorelays , the Hpt domain serves as an intermediate in the transfer of phosphoryl groups from the receiver domain phosphorylated by the HPK and the output RR . In contrast , the histidine within the Hpt domain is the initial site of phosphorylation in the chemotaxis HPK , CheA , and is phosphorylated using ATP as the phosphodonor [7] . Subsequently , the phosphoryl group is transferred from the histidine residue in the Hpt ( P1 ) domain of CheA to an aspartate residue on either of two RRs , CheY or CheB [8] . In this study , we present the structure of the Hpt ( P1 ) domain of CheA3 from R . sphaeroides in complex with its cognate RR CheY6 , which , to our knowledge , is the first structure of a Hpt domain of a CheA protein in complex with its RR . The chemotaxis pathway of Escherichia coli has been extensively characterized [9]–[12] . However , many bacteria have more complicated chemosensory pathways , employing multiple homologues of each of the chemosensory proteins [13]–[15] . R . sphaeroides has four CheA homologues and eight chemotaxis RR proteins ( six CheYs and two CheBs ) plus multiple homologues of the other E . coli chemotaxis proteins [13] , [16]–[18] . Interestingly , the different CheAs show different phosphotransfer specificities for the CheY and CheB homologues [19]–[22] . CheA3 and CheA4 form a cluster with the soluble chemoreceptors in the cytoplasm , which is believed to sense the metabolic state of the cell [13] , [23] , [24] . CheA3 and CheA4 are unusual CheAs that lack some of the domains found in E . coli CheA [20] , [25] . CheA4 has a P3 ( dimerization ) domain , a P4 ( kinase ) domain , and a P5 ( regulatory ) domain , whereas CheA3 has a P1 ( Hpt ) domain and a P5 ( regulatory ) domain separated by a 794-amino acid sequence that encodes a novel CheY-P phosphatase activity [26] . Neither CheA3 nor CheA4 is capable of autophosphorylation; instead , CheA4 phosphorylates CheA3 on H51 of the P1 ( Hpt ) domain . Once phosphorylated , CheA3-P serves as a specific phosphodonor for the chemotaxis RRs , CheY1 , CheY6 , and CheB2 . However , it is unable to phosphorylate the other chemotaxis RRs , CheY2 , CheY3 , CheY4 , CheY5 , and CheB1 [20] , [21] , [27] . This phosphotransfer specificity must be determined by the interactions between the RRs and the P1 domain of CheA3 since the isolated P1 domain of CheA3 ( CheA3P1 ) shows the same phosphotransfer specificity as full-length CheA3 [26] . In groundbreaking work , the Laub group used a mutual information bioinformatics approach to identify coevolving residues in HPKs and their cognate RRs [28] . They reasoned that many of the coevolving residues would be the specificity determinants for the phosphotransfer reaction and went on to show that it is possible to switch the RR substrate specificity of the HPK EnvZ by mutating these residues [28] . In this study , we use a structure-based approach to identify specificity determinants for the phosphotransfer reaction between CheA3P1-P and CheY6 . By introducing these residues into the noncognate RRs , we have been able to change their kinase specificity , allowing them to be phosphorylated by CheA3P1-P .
In order to elucidate the molecular details of the interaction between CheA3 and CheY6 , the complex structure of CheY6 and the unphosphorylated Hpt domain of CheA3 ( residues 1–135 ) has been solved to 1 . 40 Å using Seleno single-wavelength anomalous dispersion ( SAD ) for phasing . Data collection and refinement statistics can be found in Table 1 . CheY6 has the typical ( α/β ) 5 topology seen for E . coli CheY [29] and other structurally characterized RRs ( Figure 1 ) . Comparison with E . coli CheY reveals a high degree of structural conservation with a root mean square deviation ( rmsd ) of 1 . 4 Å over 115 Cα atoms ( 26% sequence identity ) . Although crystallized in the presence of Mn2+ , there was no additional density in the CheY6 divalent cation binding site . However , the conformation of the metal coordinating residues Asp56 , Asp9 , and Asp10 is most similar to that found in the structure of Mg2+ bound CheY [30] . Only the backbone carbonyl of Glu58 is facing away rather than pointing towards the potential metal binding site ( Figure S1 ) . CheY6 has an elongated loop region connecting β5 and α5 ( Figure 1 ) . This loop comprises 13 residues ( residues 107–119 ) in CheY6 compared to only three residues ( 109–111 ) in E . coli CheY . In the crystal , this loop is only partially ordered , suggesting that it is highly flexible . Residues 113–118 could not be traced and were omitted from the final model . The N-terminal region of this β5-α5 loop in combination with α1 of CheY6 form the vast majority of contacts to CheA3P1 ( Figure 1 ) . CheA3P1 forms a four-helix bundle ( αA–αD ) with an additional C-terminal helix ( αE ) ( Figure 1 ) . Despite low sequence identity , it is structurally very similar to previously determined CheA P1 structures [31]–[33] . Comparison with CheA P1 from Salmonella enterica serovar Typhimurium [32] and Thermotoga maritima [33] gives an rmsd of 1 . 2 Å over 116 Cα atoms ( 21% sequence identity ) and 1 . 5 Å over 101 Cα ( 18% sequence identity ) , respectively . The site of phosphorylation , His51 in CheA3P1 , is located on αB , in close proximity to the active site of the RR ( Figure 1 ) with the phosphoacceptor Asp56 on CheY6 being 7 . 5Å from His51 on CheA3P1 . Helices αA and αB face the RR and together with the αB–αC loop form the interface with CheY6 ( Figure 1 ) . To show that the conclusions drawn from the unphosphorylated structure are also valid for the physiologically relevant , phosphorylated complex , we solved the structure of phosphorylated CheA3P1 in complex with CheY6 . Since the rapid rate of phosphotransfer between CheA3P1-P and CheY6 does not allow crystallization of the wild-type , phosphorylated complex , we formed a stable , complex by introducing two substitutions , D56N and S83A , in the active site of CheY6 ( see Materials and Methods for details ) . These substitutions have previously been shown to abolish phosphotransfer from CheA3P1-P [19] . The structure was solved to 2 . 8 Å by molecular replacement using the unphosphorylated structure as model . Data collection and refinement statistics can be found in Table 1 . The structure shows clear additional density adjacent to Nε2 of His51 of CheA3P1 in agreement with phosphorylation of this residue ( Figure S2 ) . Due to the moderate resolution of the analysis , residues 60–65 , 85–97 , and 111–121 on CheY6 could not be traced and were not included in the final model . Compared to the unphosphorylated complex , CheA3P1-P undergoes a rigid body translation of 2 . 1 Å relative to the RR ( Figure 2A ) . This realignment of CheA3P1-P positions the phosphorylated His51 closer to the phosphoacceptor on CheY6 , Asp56 ( Figure 2B ) . The phosphorylated His51 is facing away from the active site of CheY6; however , a 180° flip of this side chain would put it in near linear geometry and place the phosphoryl group within 4 . 5 Å of Asp56 of CheY6 ( Figure 2B ) . The orientation of the phosphorylated His51 seen in the crystal structure is likely caused by the lack of a divalent cation in the metal binding site of the RR leading to electrostatic repulsion from the CheY6 active site . The difference in binding affinity for CheY6 between CheA3 and CheA3-P is not known . In E . coli , the difference in Kd values for CheY between CheA and CheA-P is relatively small , with both values being in the low micromolar range [34] . Consistent with this , we find very little change in the total buried surface area between the phosphorylated ( 605 Å2 ) and unphosphorylated ( 530 Å2 ) structures of CheA3P1 complexed with CheY6 . The only difference is a hydrogen bond formed between Glu58 of CheY6 and His51 of CheA3P1 in the unphosphorylated complex that is released in the phosphorylated complex . As this is not contributing to the interface of the physiologically important complex , it is therefore not discussed further . Because the interfaces in the phosphorylated and unphosphorylated complex are highly similar , the following discussion of the interactions between CheA3P1 and CheY6 are based on the high-resolution , unphosphorylated structure . The interface between CheA3P1 and CheY6 is dominated by hydrophobic interactions with only one hydrogen bond and no salt bridges formed between the two proteins . The buried surface area of the interface is small ( 530 Å2 ) , indicative of a weak interaction , consistent with the transient nature of the complexes [35] . A continuous interface is formed by two sites of interaction complemented by a hydrogen bond formed between Ser83 on CheY6 and Arg58 on CheA3P1 . One of these sites lies between the elongated loop region between β5 and α5 in CheY6 and αB and the following loop connecting αB and αC on CheA3P1 ( Figure 1A ) . The other site lies between the N-terminal end of α1 on CheY6 and αA/αB on CheA3P1 ( Figure 1B ) . The latter is the major binding site including ∼70% of the total buried surface area ( 358 Å2 ) . In this region , helices αA and αB of CheA3P1 form a hydrophobic pocket comprising Ile11 , Leu14 , and Tyr15 on αA , and Asn56 , Val59 , and Leu60 on αB ( Figure 3A and 3B ) . This pocket is situated adjacent to α1 on CheY6 and in ideal position to accommodate Met13 on the N-terminal region of this helix . This Met finger is protruding from α1 , and its side chain fits snugly into the hydrophobic pocket , with 99 . 7% of its accessible surface area being buried ( Figure 3B ) . The hydrophobic interaction is complemented by Ala12 , Leu16 , and Tyr17 of α1 on CheY6 , all facing the same side on CheA3P1 as Met13 ( Figure 3A ) . Ala12 of CheY6 extends the hydrophobic interface by interaction with Leu14 of CheA3P1 , whereas Leu16 of CheY6 interacts with Ile11 and the aliphatic part of Glu10 on αA of CheA3P1 . Tyr17 extends the binding surface towards the β5-α5 loop by interacting with Val59 at the C-terminus of αB . The elongated loop between β5 and α5 of CheY6 forms van der Waals interactions with Ser109 , Gly110 , and Thr111 of CheY6 contacting Ile59 , Gly61 , and Ser63 on αB and the following loop region on CheA3P1 ( Figure 3C ) . A main chain hydrogen bond is formed between Gly110 and Val59 . Of these interactions , the Met13 finger has the largest contribution towards the binding interface on CheY6 , accounting for almost a third ( 156 Å2 ) of the total buried surface area . Together with Ala12 , Leu16 , and Tyr17 , it is accounting for over 60% ( 321 Å2 ) of the interface . On CheA3P1 , the residues involved in the interaction with CheY6 are slightly less clustered and spatially farther apart . Yet the three residues with the biggest contribution towards the total buried surface area , Val59 ( 103 Å2 , 20% of total buried surface area ) , Leu14 ( 64 Å2 , 12% ) , and Ile11 ( 48 Å2 , 9% ) , together account for almost half of the overall binding surface . This analysis shows that only a small number of residues are necessary to form the majority of interactions within the complex . On CheY6 , these are all located on α1 . Within this helix , Met13 has the most crucial role in the hydrophobic interaction . A sequence alignment of all identified RRs in R . sphaeroides shows that only CheY6 and CheB2 have a Met residue at position 13 , whereas all others either have a Ser , Thr , or Ala at this position ( Figure 4 ) . As CheA3-P serves as a phosphodonor for both CheY6 and CheB2 this suggests that Met13 has an important role in determining the specificity for binding to CheA3P1 . Residues 11–17 of CheY6 collectively account for ∼67% of the buried surface area of CheY6 in the CheA3P1-CheY6 complex , with M13 , L16 , and Y17 each contributing ∼30% , ∼16% , and ∼8% , respectively . To confirm that this surface of CheY6 is involved in the interactions with CheA3P1-P that lead to phosphotransfer in solution , we substituted residues at this surface to mimic those found in the noncognate RRs , CheY3 and CheY4 . M13 was changed to S , and Y17 was changed to M as found in both CheY3 and CheY4 . CheY3 and CheY4 both have C at position 16; however , it has previously been shown that CheY3 ( C16S ) and CheY4 ( C16S ) behave indistinguishably from CheY3 and CheY4 in phosphotransfer assays ( S . L . Porter , unpublished data ) , and therefore , to avoid any potential problems from disulphide bond formation during the purification , we changed L16 to S rather than C . Two mutant proteins were produced , CheY6 ( M13S ) and CheY6 ( M13S , L16S , Y17M ) . Surface plasmon resonance ( SPR ) assays showed that wild-type CheY6 binds to CheA3P1 with an affinity of 218 µM ( Figure 5 ) . This weak interaction is in agreement with the transient nature of the complex and the small buried surface area between the two proteins as observed in the crystal structure . Both the single-mutant protein , CheY6 ( M13S ) , and the triple-mutant protein CheY6 ( M13S , L16S , Y17M ) showed a remarkable decrease in affinity , with a Kd of more than 1 mM ( Figure 5 and Figure S3 ) . Consistent with the SPR binding assays , the rate of phosphotransfer from CheA3P1-P to each of the two mutant CheY6 proteins was much slower than to wild-type CheY6 ( Figure 6A–6C ) , with phosphotransfer to the triple-mutant protein CheY6 ( M13S , L16S , Y17M ) being slowest . These results underline the conclusions drawn from the structural characterization of the complex and stress the importance of the Met finger at position 13 for binding of CheY6 to CheA3P1 . In addition , the hydrophobic interaction mediated by Leu16 and Tyr17 also adds significantly to recognition of CheY6 by CheA3P1 . The sequence of CheY4 is 36% identical to CheY6; however , unlike CheY6 , CheY4 is not a cognate RR of CheA3P1-P ( Figure 6D ) . Consistent with this , SPR assays failed to detect a significant interaction between CheA3P1 and CheY4 ( Figure 5 ) . In an attempt to alter the phosphotransfer specificity of CheY4 to allow phosphotransfer from CheA3P1-P , we substituted the CheY4 residues corresponding to the positions shown to interact in the CheA3P1-CheY6 structure , so that they matched CheY6 . Two mutant proteins were produced: CheY4 ( P12A , S13M ) and CheY4 ( P12A , S13M , C16L , M17Y ) . Ala12 was included in both mutant proteins since CheY4 has a Pro at position 12 that might influence the orientation of α1 and thus interfere with the proper positioning of Met13 for interaction with the hydrophobic pocket . SPR assays showed that both mutant proteins bind CheA3P1 ( Figure 5 and Figure S3 ) , and in phosphotransfer assays , both mutant proteins were phosphorylated by CheA3P1-P ( Figure 6E and 6F ) . Phosphotransfer was fastest from CheA3P1-P to CheY4 ( P12A , S13M ) , with most of the initial CheA3P1-P dephosphorylated within 10 s ( Figure 6E ) . Interestingly , although the rate of phosphotransfer from CheA3P1-P to CheY4 ( P12A , S13M ) was faster than to CheY4 ( P12A , S13M , C16L , M17Y ) , the SPR results show that CheA3P1 interacts slightly more strongly with CheY4 ( P12A , S13M , C16L , M17Y ) than with CheY4 ( P12A , S13M ) ( Figure 5 and Figure S3 ) . This apparent discrepancy could be explained by the alignment of the phosphorylated histidine and the phosphorylatable aspartate in the CheY4 ( P12A , S13M , C16L , M17Y ) . CheA3P1-P complex , which might be slightly less optimal for catalysis than in the CheY4 ( P12A , S13M ) . CheA3P1-P complex . Nevertheless , both methods show that mutating CheY4 so that it resembles CheY6 at the contact sites with CheA3P1 enhances both binding affinity and phosphotransfer rate . These results demonstrate that substitution of just two residues is sufficient to change the phosphotransfer specificity of CheY4 . Having successfully reengineered the phosphotransfer specificity of CheY4 by introducing A12 and M13 , we used the same approach to change the specificity of CheY1 , CheY3 , CheY5 , and E . coli CheY ( Figure 7 ) . These proteins share between 30% and 33% sequence identity with CheY6 . In all cases , the mutant proteins containing the alanine and methionine substitutions were phosphorylated more rapidly by CheA3P1-P than were their wild-type counterparts . Although CheY1 and E . coli CheY both lack a methionine residue at the position corresponding to M13 of CheY6 , they are both phosphorylated by CheA3P1-P ( Figure 7A and 7C ) ; however , phosphotransfer to their corresponding alanine and methionine substitution mutant proteins proceeded even faster , with almost complete dephosphorylation of CheA3P1-P within 10 s ( Figure 7B and 7D ) . Similar to CheY4 , wild-type CheY3 and CheY5 were not phosphorylated by CheA3P1-P ( Figure 7E and 7G ) , and the effect of the alanine and methionine substitutions was to allow rapid phosphotransfer from CheA3P1-P ( Figure 7F and 7H ) . These results demonstrate that the RR residues at the positions equivalent to A12 and M13 of CheY6 play a major role in determining phosphotransfer specificity .
The structure of CheA3P1 in complex with CheY6 is , to our knowledge , the first structure of a Hpt domain of a CheA protein in complex with a RR . To date , there are only two other structures of a Hpt domain in complex with a RR , namely Spo0B/Spo0F [38] , [39] from Bacillus subtilis and YPD1/SLN1 [40] , [41] from Saccharomyces cerevisiae . Although the overall structure of the RR is well conserved in all three complexes , there are remarkable differences in the Hpt domains . Spo0B exists as a dimer in which the two α-helices of the helical hairpin domain of each protomer associate to form a four-helix bundle . In addition , Spo0B has a C-terminal domain with an α/β fold , which is involved in binding of the RR . YPD1 is a monomeric Hpt protein consisting of a four-helix bundle with an additional short helix at the N-terminus . This N-terminal helix is involved in RR binding and complements the interface formed by helices αB and αC . When the three complexes were superimposed with respect to the receiver domains , CheA3P1 aligned well with YPD1 but only poorly with Spo0B ( Figure S4 ) . CheA3P1 . CheY6 shows the smallest binding interface amongst the three complexes solved so far , with 530 Å2 compared to 953 Å2 in YPD1/SLN1 and 1 , 200 Å2 in Spo0F/Spo0B . The interface is smaller since CheA3P1 has neither the C-terminal α/β fold domain seen in Spo0B nor the additional N-terminal helix seen in YPD1 . Despite this small interface , CheA3P1 shows binding and efficient phosphotransfer to CheY6 . It was previously proposed that the P2 domain of CheA-like proteins might be necessary for binding of the RR to CheA by increasing the binding interface [40] . However , CheA3 does not have a P2 domain , and the experimental data presented here and elsewhere [26] show that the P1 domain alone is sufficient for binding of and specific phosphotransfer to the cognate RR . Moreover , the structural and mutational analysis suggests that their interaction is mediated by very distinct residue clusters on the RR and the Hpt domain , the former being located on the N-terminal region of α1 and the loop region between β5 and α5 , and the latter being located on the N- and C-terminal region of αA and αB , respectively , and the loop region connecting αB and αC . This is in good agreement with a recently published computational analysis of amino acid coevolution of cognate histidine kinase–RR pairs [28] and the binding site found in the YPD1/SLN1 [40] , [41] or Spo0B/Spo0F [38] , [39] complex . The specificity of protein–protein interactions is essential for most cellular processes . Despite the vast number of these interactions , our understanding of the molecular basis of their specificity is limited . The structure presented here has allowed us to identify specificity determinants for the CheA–CheY interaction . We successfully used this information to redesign noncognate RRs to allow them to be rapidly phosphorylated by CheA3P1-P ( Figures 6 and 7 ) . Whereas the Laub group have reengineered a HPK to specifically phosphorylate non-cognate RR substrates [28] , we have now shown that it is possible to rationally reengineer a RR so that it can be phosphorylated by a noncognate HPK . The changing of phosphotransfer specificity described here represents , to our knowledge , the first example of the redesign of the intracellular part of the chemotaxis pathway and provides valuable insight into how cells mediate specificity in one of the most abundant signal transduction pathways , two-component signalling . The ability to reengineer phosphotransfer specificity coupled with recent work from the Bourret group [42] , which has shown that RR autodephosphorylation rate can be manipulated , provide a platform for the future design of synthetic two-component circuits with customizable kinetics .
CheA3P1 ( residues 1–135 of CheA3 , GenBank ID 3720125 ) , CheY6 ( GenBank ID 3720126 ) , and CheY4 ( GenBank ID 3722004 ) were cloned into the bacterial expression vector pQE80 ( Qiagen ) , which includes an N-terminal His6-tag . Sequence verified plasmids were transformed into M15pREP4 cells ( Qiagen ) and cultivated in Terrific Broth to an absorbance at 600 nm ( A600 ) of 0 . 8 . Cultures were cooled to 20°C , induced with 0 . 25 mM IPTG , and then grown for ∼15 h before harvesting . The bacterial pellets were resuspended in 25 mM sodium phosphate ( pH 8 . 0 ) , 500 mM NaCl , 0 . 5 mM β-mercaptoethanol , and EDTA-free protease inhibitor cocktail ( Roche ) . Cells were lysed using a Basic Z model cell disruptor ( Constant Systems ) and fractionated by centrifugation ( 45 , 000g , 4°C , 60 min ) . CheA3P1 , CheY6 , and CheY4 were purified from the supernatant by immobilized metal-affinity chromatography [19]–[21] . The samples were dialysed against standard buffer ( 30 mM Hepes [pH 7 . 5] , 150 mM NaCl , 5 mM sodium acetate , 2 mM manganese chloride , 2 mM Tris ( 2-carboxyethyl ) phosphine [TCEP] ) and further purified by size-exclusion chromatography . Protein purity and concentration was measured as described [43] . Purified proteins were stored at −80°C . Seleno-methionine ( SeMet ) -labelled CheY6 was produced essentially as described [44] . The complex between CheY6 and CheA3P1 was formed by mixing equimolar amounts of both proteins . The sample was incubated for 30 min at room temperature ( 20°C ) , and the complex was purified by size-exclusion chromatography . CheA3P1 was phosphorylated using ATP and CheA4 , and purified as described previously [26] . The final preparation of CheA3P1-P was free of ATP and CheA4 . Phosphorylation was verified by mass spectrometry . Equimolar amounts of phosphorylated CheA3P1 and CheY6 ( D56N , S83A ) were mixed and set up for crystallization . Mutations were introduced using the Quikchange Site-Directed Mutagenesis Kit ( Stratagene ) . Crystal trials were set up with the purified CheA3P1 . CheY6 complex at a concentration of 20 mg/ml . We set up nanoliter crystallization trials using a Cartesian Technologies robot ( 100 nl of protein solution plus 100 nl of reservoir solution ) in 96-well Greiner plates [45] , placed them in a TAP ( The Automation Partnership ) Homebase storage vault maintained at 295 K , and imaged them via a Veeco visualization system . Crystals of the phosphorylated , unphosphorylated and SeMet-labelled complex between CheY6 and CheA3P1 could be obtained in 100 mM Bicine ( pH 9 ) , 1 M LiCl , 20% ( w/v ) PEG6000 . Crystals were optimized using a dilution series of the initial condition . Crystals reached their final size of 400×100×100 µm3 after 12 h . Diffraction data were collected at 100 K , crystals being flash-cooled in a cryo N2 gas stream . Before flash-freezing , crystals were cryo-protected with perfluoropolyether oil PFO-X125/03 ( Lancaster Synthesis ) . The native dataset was collected at beamline I03 ( λ = 0 . 9757 Å ) and the SeMet dataset at beamline I04 ( λ = 0 . 979 Å ) at the Diamond Light Source . Data for the phosphorylated complex were collected on a MAR345 detector ( Marresearch ) mounted on a Miromax007 generator ( Rigaku ) , equipped with Varimax HR Osmic mirrors ( Rigaku ) . X-ray data were processed and scaled with the HKL suite [46] . Data collection statistics are shown in Table 1 . The CheA3P1-CheY6 complex structure was determined by SAD analysis . The positions of 12 selenium atoms were determined by using SHELXD [47] . This solution was put into AUTOSHARP [48] for phase calculation , improvement , and phase extension using the high-resolution native data to 1 . 4 Å resolution . The resulting map was of high quality and allowed tracing of the whole polypeptide chain . An initial model was built automatically using Arp/wARP [49] and manually adjusted using COOT [50] . The structure was refined using autoBUSTER [51] , REFMAC [52] , and Phenix [53] . The phosphorylated complex was solved by molecular replacement using Phaser [54] with the unphosphorylated complex as model . Refinement statistics are given in Table 1; all data within the indicated resolution range were included . Stereochemical properties were assessed by MOLPROBITY [55] and PROCHECK [56] . Ramachandran statistics are as follows ( favored/disallowed , % ) : CheA3P1-CheY6 unphosphorylated 98 . 8/0 , CheA3P1-CheY6 phosphorylated 99 . 1/0 . Superpositions were calculated using lsqkab [57] implemented in the CCP4 suite and electrostatic potentials were generated using APBS [58] . Buried surface areas of protein–protein interactions were calculated using the PISA Webserver ( http://www . ebi . ac . uk/msd-srv/prot_int/pistart . html ) . Coordinates are deposited in RSCB Data Bank under 3KYI and 3KYJ . SPR experiments were performed using a Biacore T100 machine ( GE Healthcare ) at 25°C in standard buffer supplemented with 0 . 05% ( v/v ) Tween 20 . Protein concentrations were determined from the absorbance at 280 nm using calculated molar extinction coefficients . Proteins for surface attachment were enzymatically biotinylated within an engineered C-terminal tag . These proteins were then attached to surfaces on which 3 , 000 response units ( RU ) of streptavidin were coupled via primary amines [59] yielding a density of 200–1 , 500 RU of biotinylated protein . All experiments were done in duplicates with independently purified proteins . The signal from experimental flow cells was corrected by subtraction of a blank and reference signal from a mock or irrelevant protein-coupled flow cell . In all experiments analyzed , the experimental trace returned to baseline after each injection and the data fitted to a simple 1∶1 Langmuir model of binding . Kd values were obtained by nonlinear curve fitting of the Langmuir binding isotherm ( bound = Cmax/ ( Kd + C ) , where C is analyte concentration and max is the maximum analyte binding ) evaluated using the Biacore Evaluation software ( GE Healthcare ) . Assays were performed at 20°C in TGMNKD buffer ( 50 mM Tris HCl , 10% [v/v] glycerol , 5 mM MgCl2 , 150 mM NaCl , 50 mM KCl , 1 mM DTT [pH 8 . 0] ) . CheA3P1 was phosphorylated using [γ-32P] ATP and CheA4 , and purified as described previously [26] . The final preparation of CheA3P1-32P was free of ATP and CheA4 . CheA3P1-32P ( 1 µM ) was mixed with 5 µM RR in a final reaction volume of 100 µl . Following the addition of RR , reaction aliquots of 10 µl were taken at the indicated time points and quenched immediately in 20 µl of 1 . 5× SDS-PAGE loading dye ( 3 . 75% [w/v] SDS , 45 mM EDTA , 18 . 75 mM Tris HCl , 18 . 75% [v/v] glycerol , 1 . 5% [v/v] β-mercaptoethanol [pH 6 . 8] ) . Quenched samples were analyzed using SDS-PAGE and phosphorimaging as described previously [21] .
|
The ability to respond to environmental stimuli is a universal feature of living cells . Evolution has created a vast array of signalling mechanisms that enable cells to react in many ways to extracellular changes . In bacteria , two-component signalling mechanisms , comprising a sensor protein kinase paired with its a cognate response regulator , are used widely to sense and respond to environmental changes . Some species of bacteria have over 150 different two-component pairs in a single cell , so the specificity between these pairs has to be tightly controlled to prevent “crossed wires” between signalling pathways . In this study , we have identified the determinants of this specificity in a two-component complex that controls the movement of Rhodobacter sphaeroides along a chemical gradient . By solving and analysing the crystal structure of this complex , we were able to pinpoint the amino acid residues that are crucially involved in formation of the complex . Knowledge of these crucial residues allowed us to convert noncognate response regulators into cognate response regulators simply by changing two amino acids . This reengineering of two-component signalling pathways paves the way for producing custom-designed circuits for applications in synthetic biology .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry/cell",
"signaling",
"and",
"trafficking",
"structures",
"cell",
"biology/cell",
"signaling"
] |
2010
|
Using Structural Information to Change the Phosphotransfer Specificity of a Two-Component Chemotaxis Signalling Complex
|
During infection by invasive bacteria , epithelial cells contribute to innate immunity via the local secretion of inflammatory cytokines . These are directly produced by infected cells or by uninfected bystanders via connexin-dependent cell-cell communication . However , the cellular pathways underlying this process remain largely unknown . Here we perform a genome-wide RNA interference screen and identify TIFA and TRAF6 as central players of Shigella flexneri and Salmonella typhimurium-induced interleukin-8 expression . We show that threonine 9 and the forkhead-associated domain of TIFA are necessary for the oligomerization of TIFA in both infected and bystander cells . Subsequently , this process triggers TRAF6 oligomerization and NF-κB activation . We demonstrate that TIFA/TRAF6-dependent cytokine expression is induced by the bacterial metabolite heptose-1 , 7-bisphosphate ( HBP ) . In addition , we identify alpha-kinase 1 ( ALPK1 ) as the critical kinase responsible for TIFA oligomerization and IL-8 expression in response to infection with S . flexneri and S . typhimurium but also to Neisseria meningitidis . Altogether , these results clearly show that ALPK1 is a master regulator of innate immunity against both invasive and extracellular gram-negative bacteria .
Intestinal epithelial cells ( IECs ) are not considered to be professional immune cells . However , they play an important role in immuno-surveillance and contribute to the initial phase of inflammation after infection by invasive bacteria or viruses . They can sense the presence of pathogens and orchestrate , together with resident macrophages , the recruitment of immune cells to sites of infection . IECs sense highly conserved pathogen-associated molecular patterns ( PAMPs ) via pathogen recognition receptors ( PRRs ) including Toll-like ( TLRs ) and NOD-like receptors ( NLRs ) . They also detect cellular stress-induced danger-associated molecular patterns ( DAMPs ) produced during infection . All these sensing mechanisms result in complex signal transduction cascades regulating the expression of proinflammatory genes coding for cytokines , chemokines and antimicrobial peptides . Shigella flexneri is an enteroinvasive bacterium responsible for shigellosis , an acute intestinal inflammation in humans [1] . After ingestion of contaminated food or water , bacteria reach the large intestine and cross the intestinal barrier by transcytosis through M-cells . Once in the submucosal area , they utilize a type III secretion ( T3S ) apparatus to induce apoptosis in macrophages and invade IECs from their basolateral side . A T3S apparatus is a syringe-like nanodevice enabling the injection of bacterial effector proteins into target cells [2] . Once effectors have translocated into cells , they can subvert the cellular activities of central host factors to favor bacterial internalization . Shigella bacteria then escape the internalization vacuole , multiply within the cytoplasm and use actin-based motility to spread from cell-to-cell within the intestinal epithelium . It has been proposed that the main PRR involved in the direct recognition of S . flexneri is the NLR NOD1 [3] . This receptor recognizes a component of the peptidoglycan called D-glutamyl-meso-diaminopimelic acid that is part of the gram-negative bacterial cell wall [4] . Upon recognition , NOD1 oligomerises and interacts with the receptor-interacting serine/threonine-protein kinase 2 ( RIP2 ) [5] . This protein associates with the transforming growth factor ( TGF ) -β-activated kinase 1 ( TAK1 ) , and the TAK1 binding protein 1 and 2 ( TAB1 and 2 ) complex . This process leads to the phosphorylation , ubiquitination and degradation of the inhibitory κB ( IκB ) , the nuclear translocation of the NF-κB transcription factor and the transcription of pro-inflammatory genes including the gene coding for interleukin-8 ( IL-8 ) . TAK1 is also involved in the activation of the MAPKs JNK , p38 and ERK , which are important for the activation of the transcription factor AP1 [6] and histone H3 phosphorylation . In addition , S . flexneri infection can also be sensed indirectly via the production of DAMPs . For instance , Dupont et al . found that the membrane vacuolar remnants produced after vacuolar lysis are detected by host cells and that the signals produced contribute to inflammation [7] . In particular , the accumulation of diacylglycerol around the bacterial entry site and within membrane remnants activates NF-κB via a mechanism dependent on the CARD–BCL10–MALT1 complex and TRAF6 [8] . Interestingly , S . flexneri possesses a number of tools downregulating the immune response of infected cells . In particular , several type III effectors interfere with the NF-κB and MAPK pathways to reduce IL-8 expression . For instance , OspG reduces the nuclear translocation of NF-κB by preventing IκB ubiquitination and degradation [9] . OspF reduces transcription via its phosphothreonine lyase activity towards p38 and ERK1/2 and its subsequent impact on chromatin remodeling [10] . Although bacteria manipulate the inflammatory response of infected cells , a massive influx of polymorphonuclear cells is observed in tissues infected with S . flexneri [11] . ATP , released by intestinal epithelial cells after infection by S . flexneri , contributes to this inflammation [12] . In addition , a previous study by our laboratory showed that innate immunity during S . flexneri infection is potentiated by a gap junction-mediated mechanism of cell-cell communication between adjacent epithelial cells [13] . We observed NF-κB and MAP kinase activation in uninfected cells located in the proximity of cells containing bacteria and showed that these bystander cells produced large amounts of inflammatory cytokines including IL-8 and tumor necrosis factor alpha ( TNFα ) . IL-8 was also largely produced in bystander cells after infection with Salmonella typhimurium and Listeria monocytogenes [13 , 14] , suggesting that potentiation of innate immunity by cell-cell communication is a common host response to different bacterial infections . This phenomenon also occurs during viral infections . First , Patel et al . found that recognition of viral double stranded DNA leads to type I interferon expression in bystander cells via a gap junction-mediated mechanism [15] . More recently , it has been shown that anti-viral immunity can spread via the diffusion of cGMP-AMP through gap junctions; cGMP-AMP then binds to the receptor STING localized at the endoplasmic reticulum , which subsequently induces anti-viral gene expression [16] . Although the control of innate immunity has important physiological consequences during bacterial infection , the molecular basis of its regulation remains poorly understood . Here we performed a genome-wide RNAi screen and identified the proteins TIFA and TRAF6 as critical factors for the control of IL-8 expression during S . flexneri infection . We show that threonine 9 ( T9 ) and the forkhead-associated domain ( FHA domain ) of TIFA are both important for the oligomerization of TIFA occurring in infected and bystander cells . This process is required for the subsequent oligomerization of TRAF6 and the activation of NF-κB . We demonstrate that TIFA/TRAF6-dependent IL-8 expression is triggered by the bacterial metabolite heptose-1 , 7-bisphosphate ( HBP ) . In addition , we identify alpha-kinase 1 ( ALPK1 ) as the critical kinase controlling TIFA oligomerization and show that ALPK1 controls innate immunity in response to the invasive bacteria S . flexneri and S . typhimurium as well as to the extracellular pathogen Neisseria meningitidis .
In order to characterize the signaling pathways controlling inflammation during infection of epithelial cells by enteroinvasive bacteria , we systematically searched for proteins regulating IL-8 expression following S . flexneri infection . For this purpose , we developed a high throughput assay that monitors IL-8 expression at the single-cell level using fluorescence microscopy ( Fig 1A ) and performed a genome-wide RNAi screen . HeLa cells , an epithelial cell line commonly used in S . flexneri infection assays , were infected for 3 . 5 hours with the ΔvirG mutant of S . flexneri as previously described [17] . This mutant is unable to perform actin-based motility [18] and forms large intracellular microcolonies , which are easily detectable by automated image analysis ( Fig 1B and S1 Fig ) . Background signals from remaining extracellular bacteria were minimized by engineering S . flexneri to express the dsRed protein only once it is intracellular [19] . dsRed expression was restricted to cytosolic bacteria by placing dsRed under the transcriptional control of the glucose 6-phosphate transporter uhpt promoter , which is only upregulated once bacteria are in the presence of glucose 6-phosphate [20] . Cells were then treated with monensin to trap IL-8 in intracellular compartments . After fixation , cells were stained for DNA , F-actin and IL-8 and visualized by immunofluorescence . In agreement with previous work [13] , IL-8 expression was largely restricted to uninfected cells located in the proximity of infected cells ( Fig 1B and S1 Fig ) , confirming the importance of bystander cell activation in the control of inflammation during S . flexneri infection [13 , 21] . In order to identify proteins involved in the control of IL-8 expression , the assay was run in a high throughput setup to screen a commercially available genome-wide library made up of pools of 4 siRNAs per gene . Total cell number , infection rates and IL-8 measurements were extracted for all targeted genes using CellProfiler ( see Materials and Methods , S1 Table ) . As expected from previous work [22] [23] , pools targeting NF-κB p65 and TAK1 had strong inhibitory effects on IL-8 expression ( S1 Table ) , validating the approach and the experimental setup of the screen . TIFA and TRAF6 were found amongst proteins whose depletion strongly inhibited IL-8 expression after S . flexneri infection , and were thus selected for further validation and molecular characterization ( Fig 1C , S1 Table ) . TRAF6 mediates signaling from members of the TNF receptor superfamily as well as the Toll/IL-1 family [24] . Interestingly , a previous publication had already reported that TRAF6 was involved in the activation of NF-κB in S . flexneri-infected cells [8] . TIFA is a 20-kDa protein that was first identified as a TRAF6-interacting protein in a yeast two-hybrid screen [25] . It contains a FHA domain , known to bind phosphothreonines and phosphoserines , and a consensus TRAF6-binding motif [26] . In TNFα signaling , it is involved in the oligomerization of TRAF6 , which is required for downstream activation of NF-κB [27] . Very recently , it has been reported that TIFA is involved in the inflammatory response triggered following the detection of heptose-1 , 7-bisphosphate ( HBP ) , a metabolite present in gram-negative bacteria [28] . HBP can be secreted or released upon bacterial lysis and internalized by eukaryotic cells via endocytosis . In order to exclude possible off-target effects in the RNAi screening data and confirm the specific implication of TIFA and TRAF6 during S . flexneri infection , silencing of these two genes was repeated with an independent set of siRNA sequences . While infection remained comparable ( S2 Fig ) , this independent approach confirmed a dramatic inhibition of IL-8 after S . flexneri ΔvirG infection of cells depleted for TIFA and TRAF6 ( Fig 1D and 1E ) . Similar results were obtained upon infection with wild-type bacteria ( Fig 1F and 1G ) as well as in HEK293 cells ( Fig 1H and 1I ) , showing that the contribution of TIFA and TRAF6 was not restricted to infections with the ΔvirG mutant or with HeLa cells . Altogether , these data show that TIFA and TRAF6 play an essential role in the control of inflammation in S . flexneri infection of epithelial cells and confirm that RNAi screens are valuable tools to identify new players in a given cellular pathway . Since a published report indicated that TRAF6 was involved in the activation of NF-κB in S . flexneri-infected cells [8] , we tested whether TIFA was also required for this process . The activation of NF-κB was monitored by following the nuclear translocation of the p65 subunit in conditions where nearly all cells were infected with S . flexneri . Interestingly , p65 translocation was reduced both in TRAF6 and TIFA-depleted cells ( Fig 2A and 2B ) , showing that these proteins were required to activate NF-κB in infected cells . When cells were infected at a lower MOI ( Fig 2C and 2D ) , a reduction of NF-κB translocation was also found in bystander cells , showing that the depletion of TIFA and TRAF6 had an impact on NF-κB activation in both cell types . The role of TIFA in NF-κB activation was more broadly tested using stimuli other than S . flexneri infection . In contrast to TRAF6 , depletion of TIFA failed to inhibit NF-κB activation induced by phorbol 12-myristate 13-acetate ( PMA ) ( Fig 2E ) , showing that TIFA is not systematically involved in pathways activating NF-κB and that TRAF6 can also function independently of TIFA . Depleting TIFA and TRAF6 had no significant effect on TNFα-induced NF-κB activation ( Fig 2F ) but partially inhibited activation induced by the NOD1 ligand C12-iE-DAP ( Fig 2G ) . Together , these results show that TIFA is not involved in the intrinsic machinery of NF-κB activation . Instead , we found TIFA to be implicated in at least two signaling pathways that link bacterial infection to inflammation . TIFA contains a FHA domain ( Fig 3A ) , a widespread signaling unit that recognizes phosphorylated threonine and serine residues and binds proteins intra- and inter-molecularly [29] . Huang et al . showed that when TIFA is unphosphorylated at the threonine 9 position , it exists as an intrinsic dimer [27] . Upon TNFα stimulation , T9 is phosphorylated by an unknown kinase and FHA-pT9 binding occurs between different dimers forming large TIFA oligomers . This mechanism leads to the subsequent oligomerization of TRAF6 and activation of NF-κB . In order to characterize the mode of action of TIFA during S . flexneri infection , we investigated the contribution of T9 and the FHA domain . For this purpose , we measured IL-8 expression after infection of cells that were first depleted for TIFA by RNAi and then transfected with siRNA-resistant wild-type or mutated TIFA cDNA constructs . As expected , we found that wild-type TIFA was able to significantly rescue IL-8 expression ( Fig 3B and 3C ) . In contrast , TIFA mutated at T9 ( T9A mutant ) or within the FHA domain ( RKN mutant ) failed to restore IL-8 expression . The same result was observed with the TIFA E178A mutant [27] , which is unable to bind TRAF6 ( Fig 3B and 3C ) . Altogether , these results show that T9 , the FHA domain and E178 are all essential for TIFA activity suggesting that , as in TNFα signaling , the pT9-FHA interaction and the ability to bind TRAF6 are necessary to induce IL-8 expression during S . flexneri infection . In order to better characterize the role of TIFA in S . flexneri infection of epithelial cells , we monitored its subcellular localization . For this , cells were transfected with a TIFA cDNA construct and TIFA was visualized after infection by immunofluorescence using a TIFA-specific antibody . In the absence of infection , the protein was uniformly distributed in the cytoplasm and the nucleus ( Fig 4A ) . Following infection with S . flexneri , punctate structures , likely corresponding to large TIFA protein oligomers [27] , were formed . These structures were still visible in S . flexneri-challenged cells after several hours ( Fig 4A and 4B ) . TIFA oligomers were found in both infected and bystander cells , suggesting that TIFA was functionally active in both cell types during infection . A co-staining between TIFA and NF-κB p65 showed that TIFA oligomers formed as early as 15 minutes post-infection and seemed to even precede NF-κB activation as visible in some cells ( Fig 4C ) . TIFA oligomerization was also observed following infection of the Caco-2 cell line ( S3 Fig ) , revealing that this process is also a relevant host response to S . flexneri infection in human colonic cells . The role of the FHA-pT9 interaction and TRAF6 binding in the mechanism of TIFA oligomerization was investigated in cells transfected with the different TIFA mutants . Neither the T9A nor the RKN mutant was able to form oligomers ( Fig 4D ) , indicating that the FHA-pT9 interaction was necessary . In contrast , the E178A mutant formed oligomers ( Fig 4D ) , demonstrating that binding to TRAF6 was not required for TIFA oligomerization . Extrapolating these data to the IL-8 rescue experiment ( Fig 3B and 3C ) suggests that TIFA oligomerization and binding to TRAF6 are both required to induce IL-8 expression after S . flexneri infection . These results further suggested that , in line with published data on TNFα signaling [27] , TIFA also induces the oligomerization of TRAF6 and the subsequent activation of NF-κB following S . flexneri infection . This hypothesis was tested by determining whether TIFA and TRAF6 co-localized after infection . The localization of both proteins was first visualized in S . flexneri-infected HeLa cells co-transfected with TIFA-myc and TRAF6-Flag cDNA constructs . As shown in Fig 4E , TRAF6 was also found in punctate structures both in infected and bystander cells . Furthermore , these structures were perfectly co-localized with TIFA oligomers . The same result was obtained upon infection of Caco-2 cells ( Fig 4F ) . Interestingly , the E178A TIFA mutant that is unable to bind TRAF6 did not co-localize with TRAF6 ( Fig 4E ) . The absence of TRAF6 oligomers in these cells showed that the formation of these structures was dependent on the ability of TIFA to bind TRAF6 . The interaction between TIFA and TRAF6 was further addressed by co-immunoprecipitation in cells transfected with TIFA-myc and TRAF6-Flag ( Fig 4G ) . A weak signal was detected in uninfected cells showing some TIFA-TRAF6 interaction under basal conditions whereas their interaction was strongly enhanced upon S . flexneri infection . As expected , this interaction was not observed when cells were transfected with the E178A TIFA mutant ( Fig 4G ) , confirming that TIFA and TRAF6 interact in a TIFA E178-dependent manner . Altogether , these results show that S . flexneri infection induces the formation of co-localizing TIFA and TRAF6 oligomers and that the TIFA-TRAF6 interaction depends on E178 of TIFA . To elucidate the mechanism triggering the activation of the TIFA/TRAF6 pathway , we tested whether TIFA was also involved in the induction of the IL-8 response observed after Listeria monocytogenes and Salmonella typhimurium infections . Like S . flexneri , these two enteroinvasive bacteria induce the secretion of the inflammatory cytokine IL-8 . In both cases , IL-8 expression is potentiated via cell-cell communication between adjacent epithelial cells [13] . Depletion of neither TIFA nor TRAF6 had an impact on L . monocytogenes-induced IL-8 production ( Fig 5A ) and TIFA failed to form oligomers after infection ( Fig 5B ) . In contrast , the depletion of either TIFA or TRAF6 abolished IL-8 expression after S . typhimurium infection ( Fig 5C ) , while TIFA formed oligomers in both infected and bystander cells ( Fig 5B ) . Since S . flexneri and S . typhimurium are both gram-negative , these results suggested that TIFA/TRAF6-dependent IL-8 expression was specifically triggered during gram-negative bacterial infections . We hypothesized that this innate immune response was induced by the recognition of HBP , a recently identified PAMP present in gram-negative bacteria [28] . HBP is a phosphorylated metabolic intermediate of lipopolysaccharide biosynthesis , produced from D-glycero-D-manno-heptose-7-phosphate by the HldE enzyme [28] ( S4A Fig ) . The role of HBP in the induction of IL-8 expression was directly tested by measuring IL-8 production in response to infection with a S . typhimurium mutant deleted for the hldE gene ( ΔhldE ) and which expressed the dsRed protein under the uhpT promoter . Data showed that infection with the ΔhldE mutant , which is unable to synthesize HBP , failed to induce IL-8 production both in infected and bystander cells ( Fig 5D and 5E ) . As expected , infection with bacteria deficient for the enzymes GmhB ( ΔgmhB ) or WaaC ( ΔwaaC ) , which act downstream of HldE in the ADP heptose biosynthetic pathway [30] ( S4A Fig ) , induced strong IL-8 expression ( Fig 5D and 5E , S4B Fig ) . The same experiment was repeated with S . flexneri mutants . Interestingly , the ΔhldE and ΔwaaC mutants were dramatically more invasive than wild-type or ΔgmhB bacteria ( Fig 5G and S4C Fig ) . However , at all multiplicities of infection tested , the absence of HBP led to a complete inhibition of IL-8 expression ( Fig 5F and 5H ) . As with S . typhimurium , the ΔgmhB and Δwaac mutants induced massive IL-8 expression , indicating that the expression of IL-8 was dependent on bacterial synthesis of HBP ( Fig 5F and 5H , S4D Fig ) . In addition , infection with the S . flexneri and S . typhimurium ΔhldE mutants failed to induce the oligomerization of TIFA ( Fig 5I , S5A and S5B Fig ) . Finally , multiplex cytokine analysis showed that S . flexneri infection of HeLa cells induced the secretion of IL-6 , IL-1β , IFNγ , IL-8 and TNFα in an HBP-dependent manner ( S6A Fig ) . Furthermore , the induction of IL-8 and TNFα observed in Caco-2 cells after S . flexneri infection was also largely dependent on HBP ( Fig 5J and S6B Fig ) . Altogether , these results show a causal link between HBP , the oligomerization of TIFA/TRAF6 , the activation of NF-κB and inflammatory cytokine expression . They also show for the first time , that HBP is a critical PAMP that triggers inflammation in epithelial cells during infection by at least two invasive gram-negative pathogens , S . typhimurium and S . flexneri . The observation that TIFA oligomerization was dependent on T9 and the FHA domain of TIFA suggested that at least one kinase was involved upstream of TIFA to control IL-8 expression . In order to identify kinase candidates , an RNAi screen targeting each gene of the human kinome with three individual siRNAs , was performed . TAK1 , known to be involved in S . flexneri-induced NF-κB activation downstream of TRAF6 and RIPK2 [8] , was the strongest negative hit ( S2 Table , Fig 6A ) . We tested whether this kinase could also control TIFA oligomerization during infection . Although depleting TAK1 completely abrogated IL-8 production ( Fig 6A and 6B , S2 Table ) , TIFA oligomers were still visible in infected and bystander cells ( Fig 6B ) , confirming that TAK1 was implicated downstream of TIFA . The second strongest hit was ALPK1 . ALPK1 belongs to the atypical kinase group [31] and is poorly characterized . It is a component in apical transport of epithelial cells [32] . Furthermore , polymorphism in the alpk1 gene is associated with type 2 diabetes , dyslipidemia , gout and chronic kidney disease [33–36] . Strikingly , the alpk1 and tifa genes are direct neighbors on human chromosome 4 [37] , suggesting that they may be co-regulated and part of a common cellular pathway . ALPK1 was thus further investigated for its implication in S . flexneri infection and TIFA-dependent innate immunity . First , the role of ALPK1 in IL-8 production after S . flexneri infection was confirmed by intracellular IL-8 staining ( Fig 6C ) and ELISA ( S7A Fig ) . The secretion of IL-6 , IL-1β , IFNγ and TNFα was reduced in ALPK1-depleted cells ( S7B Fig ) , showing that ALPK1 is a master regulator of S . flexneri-induced inflammatory cytokine expression , a process largely triggered in response to HBP ( S6A Fig ) . Since TIFA and TRAF6 regulated S . flexneri-induced NF-κB activation , we investigated the role of ALPK1 in this process . Western blot experiments performed on uninfected and infected cells revealed that depleting ALPK1 reduced the degradation of the inhibitor of NF-κB , IκBα , in infected cells ( Fig 6D ) . In agreement , ALPK1 depletion also impaired the nuclear translocation of NF-κB after S . flexneri infection without significantly affecting bacterial entry ( Fig 6E and 6F and S2 Fig ) . Altogether , these results suggested that ALPK1 was a promising candidate for the control of TIFA-dependent innate immunity . The role of ALPK1 in this process was directly addressed by several means . First , depletion of ALPK1 prevented the formation of TIFA oligomers both in infected and bystander cells during S . flexneri infection ( Fig 6G and 6H ) . Second , in rescue experiments , whereby cells were transfected with control or ALPK1 siRNA and then transfected with the empty vector pEYFP or a siRNA-resistant full-length ALPK1-YFP cDNA construct ( Fig 6I and 6J ) , overexpression of YFP-ALPK1 did not induce the formation of TIFA oligomers in the absence of infection , indicating that this process was tightly regulated . Notably , when ALPK1-depleted cells were transfected with full-length YFP-ALPK1 , TIFA oligomerization was restored in a large fraction of infected and bystander cells . This result excluded the possible contribution of RNAi off target effects and unambiguously established the role of ALPK1 in S . flexneri-induced TIFA oligomerization . Interestingly , transfection of a siRNA-resistant cDNA construct deleted for the kinase domain of ALPK1 ( YFP-ALPK1-ΔK ) failed to rescue TIFA oligomerization ( Fig 6I and 6J ) , showing that the kinase domain of ALPK1 was necessary for the induction of TIFA oligomerization after S . flexneri infection . Finally , the role of ALPK1 on the TIFA-TRAF6 interaction was investigated by co-immunoprecipitation experiments . Data showed that the TIFA-TRAF6 interaction induced upon S . flexneri infection was strongly reduced in ALPK1-depleted cells , demonstrating that this interaction was ALPK1-dependant ( Fig 6K ) . Altogether , these results showed that ALPK1 is a master regulator of cytokine expression during S . flexneri infection and that TIFA oligomerization depends on the kinase domain of ALPK1 . As S . flexneri-induced TIFA oligomerization occurred in response to HBP ( Fig 5I ) , we tested whether ALPK1 was involved in this process . Cells were stimulated with lysates from S . flexneri containing an empty pUC19 vector or expressing the HBP-synthesizing enzyme HldA from N . meningitidis [28] . As expected , the lysate from HldA-overexpressing bacteria was more potent at inducing IL-8 expression than those of wild-type bacteria ( Fig 7A and 7B ) . Interestingly , depletion of ALPK1 prevented the oligomerization of TIFA ( Fig 7C ) as well as IL-8 production ( Fig 7A and 7B ) in response to both lysates , showing that ALPK1 controlled the oligomerization of TIFA following HBP recognition . As with TIFA , depletion of ALPK1 failed to inhibit IL-8 expression and NF-κB activation observed after L . monocytogenes infection ( S8A and S8B Fig ) , suggesting a specific implication in infection by invasive gram-negative bacteria . Furthermore , ALPK1 was not required to activate NF-κB in response to PMA ( S9A Fig ) or TNFα ( S9B Fig ) . As with TIFA and TRAF6 , depleting ALPK1 had a moderate but significant effect on C12-iE-DAP-induced NF-κB activation ( S9C Fig ) . The role of ALPK1 was further characterized in the inflammatory response triggered by Neisseria meningitidis , an important gram-negative extracellular human pathogen . This bacterium is responsible for meningitis and other forms of meningococcal diseases including meningococcemia , a case of life-threatening sepsis [38] . Upon infection with this pathogen , HBP can be secreted or released by lysing bacteria [28] . We confirmed that treating HeLa cells with N . meningitidis lysate induced TIFA oligomerization ( Fig 7D and 7E ) and IL-8 expression ( Fig 7F ) . Furthermore , depleting either TIFA or TRAF6 prevented IL-8 expression ( Fig F ) . Interestingly , we found that TIFA oligomerization and IL-8 expression were both completely abrogated in ALPK1-depleted cells ( Fig 7D , 7E and 7F ) , showing that ALPK1 also controls the innate immune response to N . meningitidis infection ( Fig 7G ) . Altogether , these results show that HBP is a key bacterial PAMP sensed by epithelial cells during infection by both invasive and extracellular gram-negative bacteria and that TIFA/TRAF6-dependent innate immunity against HBP is controlled by ALPK1 .
An RNAi screen implicated TIFA and TRAF6 in the control of IL-8 expression after S . flexneri infection . We show that these two proteins act upstream of NF-κB p65 activation in infected and bystander cells . In particular , we provide evidence demonstrating that S . flexneri induces the oligomerization of TIFA and TRAF6 in infected and bystander cells in a FHA/T9-dependent manner . In cells expressing a TIFA mutant unable to bind TRAF6 , the formation of TRAF6 oligomers was not observed , showing that the TIFA-TRAF6 interaction is necessary to trigger TRAF6 oligomerization . Given that TRAF6 oligomerization has been shown to increase its E3 ubiquitin ligase activity [39] , our data suggest that TIFA works as an adaptor protein promoting TRAF6 oligomerization and thereby NF-κB activation and inflammatory gene expression ( Fig 7G ) . In infected and bystander cells , TIFA oligomers are distributed evenly throughout the cytoplasm . They appear within minutes of infection and are still visible four hours post infection . Co-staining of TIFA and lysosomal-associated membrane protein 1 ( LAMP1 ) in S . flexneri-infected cells revealed that TIFA/TRAF6 oligomers are not localized to lysosomes ( S10 Fig ) . More work is needed to determine whether these aggregation platforms are associated with other subcellular structures or whether they freely diffuse in cells . We show that during S . flexneri and S . typhimurium infection , the TIFA/TRAF6 pathway is activated in response to the bacterial monosaccharide HBP , present in gram-negative bacteria . Indeed , we found that the ΔhldE mutants of S . flexneri and S . typhimurium , which are unable to synthesize HBP , fail to induce the oligomerization of TIFA and the production of IL-8 . These results open up a new avenue to understand the molecular processes controlling inflammation in bacterial infection and highlight the central role of HBP during infection by invasive bacteria . In contrast to the study by Gaudet et al . [28] , the production of IL-8 in response to S . flexneri and S . typhimurium infection is unlikely due to the simple mechanism of HBP endocytosis . Indeed , we previously demonstrated that noninvasive S . flexneri bacteria failed to induce IL-8 expression [13] . This point was further confirmed by Lippmann et al . who showed that the expression of IL-8 in bystander cells requires bacterial internalization [21] . Mechanisms explaining how HBP could therefore be detected within minutes of bacterial invasion have to be envisioned . Although there is , to our knowledge , no evidence in the literature for the release of metabolites via type III secretion , we cannot exclude the possibility that HBP may be directly secreted into the host cytoplasm via the injectisome . An alternative mechanism would consist in the cellular uptake of HBP during the process of bacterial internalization . A study using dynamic imaging and advanced large volume correlative light electron microscopy recently reported that two distinct compartments are formed during the first step of bacterial invasion: the bacterial containing vacuole ( BCV ) and surrounding macropinosomes [40] . Whereas the membrane of the BCV tightly surrounds the bacterium , macropinosomes are heterogeneous in size and contain significant volumes of extracellular fluid [40] . HBP , released from residual secretion or bacterial lysis , may be engulfed by infected cells via the BCV or macropinosomes and released into the cytoplasm shortly after membrane rupture . The small molecular size of HBP ( 370 Da ) should allow its diffusion to adjacent cells via gap junctions leading to TIFA oligomerization and IL-8 expression in bystander cells ( Fig 7G ) . Alternatively , HBP sensing in infected cells may lead to the production of a second messenger that could diffuse to bystander cells and activate the ALPK1/TIFA/TRAF6 pathway . In the case of S . typhimurium , the complete rupture of the internalization vacuole is a rare event . In most cases , bacteria remain inside Salmonella-containing compartments . Interestingly , a recent study shows that early Salmonella-containing compartments are leaky and that autophagy proteins promote the repair of endosomal membranes damaged by the type III secretion system 1 [41] . In this context , HBP may leak out of these early compartments , be released into the cytoplasm of infected cells and induce IL-8 expression both in infected and bystander cells , as observed previously [13] . We showed that secretion of inflammatory cytokines after S . flexneri infection of epithelial cells in vitro is largely HBP-dependent , which supports a central role of HBP in the control of innate immunity in S . flexneri infection . More work is needed to determine the exact contribution of HBP in in vivo infection where other PAMPs , including peptidoglycan-derived peptides and LPS , have previously been shown to play a role [4 , 42] . Our results show that TIFA’s activity in S . flexneri-induced IL-8 expression is dependent on residue T9 and the FHA domain of TIFA . As the interaction between these two features occurs once T9 is phosphorylated and is required to trigger TIFA oligomerization , we searched for a kinase acting upstream of TIFA oligomerization in bacterial infection . We identified the kinase ALPK1 in a human kinome RNAi screen . Strikingly , the genes coding for ALPK1 and TIFA are immediate neighbors on human chromosome 4 [37] . Gene neighborhood is conserved across several species including coelacanth , xenopus , chicken and mouse , suggesting that both genes may be co-regulated and the encoded proteins part of a same cellular pathway . We show that depleting ALPK1 strongly reduced NF-κB activation and the production of several cytokines including IL-8 , TNFα , IL-1β , IFNγ and IL-6 after S . flexneri infection . IL-8 production was also reduced after S . typhimurium infection . ALPK1 depletion completely prevented the formation of TIFA oligomers after S . flexneri infection , a process triggered in response to HBP sensing . TIFA oligomerization was restored by overexpressing a siRNA-resistant full length ALPK1 construct . In contrast , overexpressing a construct deleted of the kinase domain of ALPK1 failed to do so , showing that the kinase domain of ALPK1 is essential for the regulation of TIFA oligomerization . In addition , co-immunoprecipitation experiments revealed that the TIFA-TRAF6 interaction is dependent on ALPK1 . All these results demonstrate that ALPK1 is involved in the early signaling cascade controlling inflammation following cellular invasion by gram-negative bacterial pathogens . Furthermore , we show that ALPK1 is also implicated in the control of inflammation after stimulation with N . meningitidis lysates , indicating that this kinase acts as a master regulator of innate immunity to both invasive and extracellular gram-negative bacteria . ALPK1 is an atypical kinase belonging to the α-kinase family that recognizes phosphorylation sites in the context of an alpha-helical conformation [31] . The fact that T9 is not in this environment is not sufficient to exclude that ALPK1 can directly phosphorylate TIFA . Indeed , it has been shown that members of this protein family can also phosphorylate substrates independently of a helical conformation [31] . More experiments are required to elucidate the mode of action of ALPK1 in the activation of the TIFA/TRAF6 pathway . In addition , it will be informative to determine whether HBP can directly bind to ALPK1 or whether this new bacterial PAMP binds to a yet unknown pathogen recognition receptor able to activate ALPK1 and trigger TIFA oligomerization . Interestingly , by sensing the presence of HBP , a metabolite of the LPS biosynthetic pathway , such a receptor would constitute a new specific sensor for the presence of gram-negative bacteria . In conclusion , we show that ALPK1 is a master regulator of innate immunity against both invasive and extracellular gram-negative bacteria . This kinase acts in response to the detection of HBP to activate the TIFA/TRAF6 pathway . By regulating the expression of inflammatory cytokines , this new signaling pathway is critical to orchestrate the initial host immune response and limit bacterial dissemination within infected tissues . It may also contribute to the control of intestinal homeostasis by regulating the molecular cross-talk taking place between gram-negative bacteria of the microbiota , the intestinal epithelium and the immune system .
HeLa ( American Type Culture Collection ) and HEK293 ( American Type Culture Collection ) cells were cultured in Dulbecco’s modified Eagle’s ( DMEM ) medium supplemented with 10% FCS and 2 mM Glutamax-1 . Caco-2 cells ( American Type Culture Collection ) were cultured in MEM , 20% FCS and 1% non-essential amino acids . Transfection of siRNAs was carried out using RNAiMAX ( Invitrogen ) . HeLa cells , seeded in 96-well plates ( 6 , 000 cells/well ) , were reverse transfected with 20 nM siRNA according to the manufacturer’s instruction . Cells were used 72 hours after transfection . siRNAs against TIFA ( s40984 ) , TRAF6 ( s14389 ) and ALPK1 ( s37074 ) were from Ambion and TAK1 from Dharmacon . For cDNA transfection , HeLa cells were seeded in a 96-well plate at a density of 12 , 500 cells/well . The next day , cells were transfected with 80 ng of plasmid using Fugene 6 ( Roche ) according to the manufacturer’s instruction . Wild-type , T9A , E178A and the RKN TIFA cDNA constructs [27] were kindly provided by Prof . M . D . Tsai ( Institute of Biological Chemistry , Academia Sinica , Taiwan ) . They were made TIFA siRNA ( s40984 ) resistant by the introduction of 3 silent point mutations within the recognition site of the siRNA . Point mutations were introduced by overlapping PCR using primers TIFA_BamHI_F , TIFA_R2 , TIFA_F2 , TIFA_XbaI_R and TIFA_EA_XbaI_R ( listed in S3 Table ) . The resulting PCR products were digested with BamHI and XbaI and ligated into pcDNA3 . A YFP-ALPK1 construct was kindly provided by Pr R . Jacob ( Marburg University , Germany ) . It was made siRNA ( s37074 ) -resistant by the introduction of 5 silent point mutations at positions 761-762-763-767-768 by directed mutagenesis ( Agilent Technology ) . A mutant deleted for the kinase domain of ALPK1 was generated by introducing a stop codon at position 3059 before the kinase domain by directed mutagenesis . All primers used in directed mutagenesis are listed in S3 Table . For TIFA and ALPK1 rescue experiments , Hela cells were first reverse transfected with TIFA or ALPK1 siRNAs ( s40984 , s37074 respectively ) . After 48 hours , they were transfected with the different siRNA-resistant TIFA cDNA constructs or siRNA-resistant full length or kinase domain-deleted YFP-ALPK1 . As a negative control , cells were transfected with the empty vectors pcDNA or pEYFP . Wild-type Flag-TRAF6 cDNA was a gift from John Kyriakis ( Addgene plasmid # 21624 ) [43] . The M90T wild-type Shigella flexneri strain and the icsA ( virG ) deletion mutant have been previously described [44] . The Salmonella typhimurium 12023 strain expressing pKD46 was provided by J . Guignot ( Institut Cochin , Paris , France ) and the EGDe . PrfA Listeria monocytogenes strain stably expressing GFP [45] was provided by Prof . P . Cossart ( Institut Pasteur , Paris , France ) . All Shigella and Salmonella strains were transformed with the pMW211 plasmid and constitutively express the dsRed protein [13] . When specifically mentioned , bacteria were alternatively transformed with a variant of pMW211 expressing dsRed under the control of the uhpT promoter ( PuhpT::dsRed ) [17] . For Neisseria meningitidis , a piliated capsulated Opc- Opa- variant of serogroup C strain 8013 named 2C43 was used . The hldA deficient mutant was obtained as previously described in [46] . S . flexneri M90T and S . typhimurium 12023 deletion mutants were generated by allelic exchange using a modified protocol of lambda red-mediated gene deletion [47] . Briefly , to obtain the S . flexneri M90T and S . typhimurium hldE ( ΔhldE ) , gmhB ( ΔgmhB ) and waaC ( ΔwaaC ) deletion mutants , the kanamycin cassette of the pkD4 plasmid was amplified by PCR with the primers listed in S3 Table . The purified PCR product was electroporated into the wild-type strains expressing the genes for lambda red recombination from the pKM208 ( for S . flexneri mutants ) or pKD46 ( for S . typhimurium mutants ) plasmids [48] . Recombinants were selected on TSB or LB plates containing 50 μg ml-1 of kanamycin . Single colonies were screened by PCR . S . flexneri M90T overexpressing the hldA gene from Neisseria meningitidis was generated as follows . The hldA gene was amplified by PCR from a bacterial lysate with the primers listed in S3 Table . After gel purification ( Macherey-Nagel ) , the PCR product was digested with EcoRI and HindIII , and ligated into EcoRI/HindIII-digested pUC19 ( Life Technology ) . The ligation product was used to transform Top10 E . Coli . pUC19-HldA was purified and used to electroporate S . flexneri M90T . As a control , S . flexneri M90T was also electroporated with the pUC19 empty vector . Bacterial lysates were prepared as described in Gaudet et al . [28] . Briefly , 1 OD600 unit of bacteria from an overnight culture was centrifuged , resuspended in 100 μl PBS and boiled for 15 mins . Bacterial debris were removed by centrifugation at 13 000 rpm for 10 mins . Supernatants were collected and protein concentration was measured by BCA assay ( Interchim ) for normalization . Lysates were then treated with RNAse A ( 10 μg/ml ) , DNAse I ( 20U ) ( both Roche ) and proteinase K ( 100 μg/ml ) ( Sigma-Aldrich ) . Samples were boiled for a further 5 minutes , centrifuged and the supernatant passed through a 0 . 22 μm filter . Lysates were stored at -20°C . S . flexneri , S . typhimurium and L . monocytogenes were used in exponential growth phase . Shigella and Salmonella were coated , or not , with poly-L-lysine prior to infection . Cells seeded in 96-well plates , were infected at indicated MOIs in DMEM supplemented with 10 mM Hepes and 2 mM glutamax-1 . After adding bacteria , plates were centrifuged for 5 minutes and placed at 37°C for indicated time periods . Extracellular bacteria were killed by gentamicin ( 100 μg/ml ) . For stimulation experiments , cells were incubated with PMA ( Sigma ) , C12-iEDAP ( Invivogen ) and TNFα ( R&D Systems ) at indicated concentrations . For intracellular IL-8 measurements , monensin ( 50 μM ) was added together with gentamycin to block IL-8 secretion . Infection and stimulation assays were stopped by 4% PFA fixation . The screening methodology has already been described [17] . Briefly , RNA interference ( RNAi ) directed against the human genome was achieved using the commercially available genome-wide siRNA library from Dharmacon ( pools of 4 siRNAs/gene ) . The human kinome RNAi screen was performed with the Ambion library made of three individual siRNAs per gene . All experiments were conducted in a 384-well plate format . In addition to gene-specific siRNAs , all plates contained general siRNA controls for transfection efficiency ( e . g . Kif11 ) , positive control siRNAs known to affect inflammation after S . flexneri infection ( TAK1 , p65 NF-κB ) and non-targeting siRNAs . In each experiment , 25 μl of RNAiMAX/DMEM ( 0 . 1 μl/24 . 9 μl ) mixture was added to each well of the screening plates containing 1 . 6 pmol siRNA diluted in 5 μl RNase-free ddH2O . Screening plates were incubated at room temperature ( RT ) for 1 hour . Following incubation , 600 HeLa CCL-2 cells were added per well in a volume of 50 μl DMEM/16% FCS , resulting in a final FCS concentration of 10% . Plates were incubated at 37°C and 5% CO2 for 72 h prior to infection . For infection , S . flexneri M90T ΔvirG pCK100 ( PuhpT::dsRed ) were harvested in exponential growth phase and coated with 0 . 005% poly-L-lysine ( Sigma-Aldrich ) . Afterwards , bacteria were washed with PBS and resuspended in assay medium ( DMEM , 2 mM L-Glutamine , 10 mM HEPES ) . 20 μl of bacterial suspension was added to each well with a final MOI of 15 . Plates were then centrifuged for 1 min at 37°C and incubated at 37°C and 5% CO2 . After 30 min of infection , 75 μl were aspirated from each well and monensin ( Sigma ) and gentamicin ( Gibco ) were added to a final concentration of 66 . 7 μM and 66 . 7 μg/ml , respectively . After a total infection time of 3 . 5 hours , cells were fixed with 4% PFA for 10 minutes . Liquid handling was performed using the Multidrop 384 ( Thermo Scientific ) for dispension steps and a plate washer ( ELx50-16 , BioTek ) for aspiration steps . For immunofluorescent staining , cells were washed with PBS using the Power Washer 384 ( Tecan ) . Subsequently , cells were incubated with a mouse anti-human IL-8 antibody ( 1:300 , BD Biosciences ) in staining solution ( 0 . 2% saponin in PBS ) for 2 hours at RT . After washing the cells with PBS , Hoechst ( 5 μg/ml , Invitrogen ) , DY-495-phalloidin ( 1 . 2 U/ml , Dyomics ) and Alexa Fluor 647-coupled goat anti-mouse IgG ( 1:400 , Invitrogen ) were added and incubated for 1 hour at RT . The staining procedure was performed using the Biomek NXP Laboratory Automation Workstation ( Beckman Coulter ) . Microscopy was performed with Molecular Devices ImageXpress microscopes . MetaXpress plate acquisition wizard with no gain , 12 bit dynamic range , 9 sites per well in a 3×3 grid with no spacing and no overlap and laser-based focusing was used . Robotic plate handling was used to load and unload plates ( Thermo Scientific ) . A 10X S Fluor objective with 0 . 45NA was used . Data analysis was performed using the computational infrastructure described in [17] . Cell counts , rates of infection and IL-8 positive cells were quantified as described in [17] . In brief , intensity and texture features were extracted from bacterial and IL-8 images . Based on these features , cells were scored for infection and IL-8 expression using CellClassifier and supervised machine learning using a Support Vector Machine based binary classifier [49] . Measurements were normalized for plate-to-plate variations and population context dependency as described in [17] . After fixation , cells were permeabilized in 0 . 1% Triton X-100 for 10 minutes , incubated in PBS supplemented with 0 . 5% BSA for 2 hours and then overnight at 4°C with different combinations of primary antibodies . NF-κB p65 localization was visualized by using a mouse monoclonal anti-p65 antibody ( Santa Cruz Biotechnology , USA ) , TIFA was visualized with a polyclonal rabbit anti-TIFA primary antibody ( Sigma-Aldrich ) , and LAMP1 was visualized with an anti-mouse anti-LAMP1 ( Abcam ) . Cells were then stained with Alexa 647- or Alexa 488-conjugated secondary antibodies ( Invitrogen , Carlsbad , USA ) . DNA and F-actin were stained with Hoechst and FITC-phalloidin , respectively . The production of IL-8 was measured by immunofluorescence using an anti-human IL-8 antibody in 0 . 2% saponin in PBS ( BD Pharmingen , San Jose , USA ) 4 hours post infection . Images were automatically acquired with an ImageXpress Micro ( Molecular devices , Sunnyvale , USA ) . Image analysis was performed using the custom module editor ( CME ) of MetaXpress . Briefly , cell nuclei were identified by the "autofind blobs" function of the CME . Nuclei were then extended by 6 pixels to define the cellular mask of each cell that was used to measure bacteria and IL-8 signals . Bacteria and IL-8 signals were both detected with the "keep marked object" function of the CME based on minimal/maximal size requirements and intensity threshold . Cells showing IL-8 signals above threshold were defined as IL-8 positive . Quantification of NF-κB activation was performed with the "translocation enhanced" module of MetaXpress ( Molecular Devices , USA ) that automatically identifies the nuclei and cytoplasmic compartments from a Hoechst image . Quantification was done by measuring the intensity ratio of p65 in the nucleus and the cytoplasm in several thousand cells per well and in three wells per condition . Cells showing nuclear/cytoplasmic p65 intensity ratio above a threshold ratio were defined as NF-κB positive cells . Cells were plated in 6-well plates ( 180 000 cells/well ) , transfected or not with 20 nM siRNA and/or 2 . 4 μg cDNA and infected according to the experiment . After infection , cells were washed twice in ice cold PBS with gentamicin ( 100 μg/ml ) , lysed in RIPA buffer supplemented with inhibitors of proteases ( Promega ) and phosphatases ( Thermofisher Scientific ) , incubated on ice for 30 minutes and subsequently centrifuged at 4°C for 30 minutes at 16 , 000g . The BCA Protein Assay kit ( Interchim ) was used to determine protein concentration . 15–20 ug of protein was subjected to SDS-polyacrylamide gels and electroblotted onto nitrocellulose membranes . For immunoprecipitation ( IP ) , cell lysates were incubated with an anti-myc antibody ( 9E10 , Santa Cruz ) overnight . Protein A/G-coated beads ( ThermoFisher ) were then added for 2 hours and washed six times in Mac Dougall buffer . Cell lysates and IPs were diluted in Laemmli buffer containing SDS and β-mercaptoethanol , boiled for 6 minutes and subjected to SDS-PAGE . Immunoblotting was performed using primary antibodies diluted in phosphate buffered saline containing 0 . 1% Tween and 5% nonfat dry milk . HRP-conjugated secondary antibodies were purchased from GE Healthcare or Cell signaling technology or ThermoFisher Scientific . The blots were developed with an enhanced chemiluminescence method ( SuperSignal West Pico Chemiluminescent substrate , Thermofisher Scientific ) . IL-8 secretion was measured by ELISA in the supernatant of HeLa and Caco-2 cells infected with S . flexneri for 6 hours . The cell-free supernatants from triplicate wells were analyzed for their IL-8 content using the commercial ELISA kit ( eBioscience ) . The secretion of additional cytokines including TNFα , IL-1β , IL-6 and IFNγ was measured using the Cytokine Human Magnetic 10-plex Panel for Luminex Platform ( Life Technologies ) . Data are expressed as mean ± standard deviation of triplicates samples as indicated . p values were calculated with a two-tailed two-sample equal variance t-test .
|
Epithelial cells line internal body cavities of multicellular organisms . They represent the first line of defense against various pathogens including bacteria and viruses . They can sense the presence of invasive pathogens and initiate the recruitment of immune cells to infected tissues via the local secretion of soluble factors , called chemokines . Although this phenomenon is essential for the development of an efficient immune response , the molecular mechanism underlying this process remains largely unknown . Here we demonstrate that the host proteins ALPK1 , TIFA and TRAF6 act sequentially to activate the transcription factor NF-κB and regulate the production of chemokines in response to infection by the pathogens Shigella flexneri , Salmonella typhimurium and Neisseria meningitidis . In addition , we show that the production of chemokines is triggered after detection of the bacterial monosaccharide heptose-1 , 7-bisphosphate , found in gram-negative bacteria . In conclusion , our study uncovers a new molecular mechanism controlling inflammation during infection by gram-negative bacteria and identifies potential targets for treatments aiming at modulating inflammation during infection .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"cytokines",
"pathology",
"and",
"laboratory",
"medicine",
"hela",
"cells",
"gene",
"regulation",
"pathogens",
"biological",
"cultures",
"immunology",
"microbiology",
"shigella",
"bacterial",
"diseases",
"developmental",
"biology",
"signs",
"and",
"symptoms",
"enterobacteriaceae",
"materials",
"science",
"molecular",
"development",
"cell",
"cultures",
"oligomers",
"bacteria",
"bacterial",
"pathogens",
"materials",
"by",
"structure",
"salmonella",
"typhimurium",
"research",
"and",
"analysis",
"methods",
"small",
"interfering",
"rnas",
"infectious",
"diseases",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"inflammation",
"medical",
"microbiology",
"gene",
"expression",
"microbial",
"pathogens",
"cell",
"lines",
"salmonella",
"immune",
"response",
"immune",
"system",
"shigella",
"flexneri",
"biochemistry",
"rna",
"cell",
"staining",
"diagnostic",
"medicine",
"nucleic",
"acids",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"cultured",
"tumor",
"cells",
"physical",
"sciences",
"non-coding",
"rna",
"organisms"
] |
2017
|
ALPK1 controls TIFA/TRAF6-dependent innate immunity against heptose-1,7-bisphosphate of gram-negative bacteria
|
Several infrequent genetic polymorphisms in the SERPINA1 gene are known to substantially reduce concentration of alpha1-antitrypsin ( AAT ) in the blood . Since low AAT serum levels fail to protect pulmonary tissue from enzymatic degradation , these polymorphisms also increase the risk for early onset chronic obstructive pulmonary disease ( COPD ) . The role of more common SERPINA1 single nucleotide polymorphisms ( SNPs ) in respiratory health remains poorly understood . We present here an agnostic investigation of genetic determinants of circulating AAT levels in a general population sample by performing a genome-wide association study ( GWAS ) in 1392 individuals of the SAPALDIA cohort . Five common SNPs , defined by showing minor allele frequencies ( MAFs ) >5% , reached genome-wide significance , all located in the SERPINA gene cluster at 14q32 . 13 . The top-ranking genotyped SNP rs4905179 was associated with an estimated effect of β = −0 . 068 g/L per minor allele ( P = 1 . 20*10−12 ) . But denser SERPINA1 locus genotyping in 5569 participants with subsequent stepwise conditional analysis , as well as exon-sequencing in a subsample ( N = 410 ) , suggested that AAT serum level is causally determined at this locus by rare ( MAF<1% ) and low-frequent ( MAF 1–5% ) variants only , in particular by the well-documented protein inhibitor S and Z ( PI S , PI Z ) variants . Replication of the association of rs4905179 with AAT serum levels in the Copenhagen City Heart Study ( N = 8273 ) was successful ( P<0 . 0001 ) , as was the replication of its synthetic nature ( the effect disappeared after adjusting for PI S and Z , P = 0 . 57 ) . Extending the analysis to lung function revealed a more complex situation . Only in individuals with severely compromised pulmonary health ( N = 397 ) , associations of common SNPs at this locus with lung function were driven by rarer PI S or Z variants . Overall , our meta-analysis of lung function in ever-smokers does not support a functional role of common SNPs in the SERPINA gene cluster in the general population .
Alpha1-antitrypsin ( AAT ) is a serum marker for inflammation produced in the liver . Its main function is to inhibit neutrophil elastase and consequently protect pulmonary tissue . The SERPINA1 gene encoding the AAT protein is known to be polymorphic in the general population . The best studied single nucleotide polymorphisms ( SNPs ) causing a reduction in AAT serum levels are the protease inhibitor S ( PI S , rs17580 ) and the protease inhibitor Z ( PI Z , rs28929474 ) variants [1] . The loss of function mechanism is especially well investigated for the PI Z variant . The resulting amino acid change in AAT leads to the protein's intracellular polymerization in hepatocytes and therefore to a reduced level of secreted serum AAT [2] . Homozygosity for PI Z ( PI ZZ genotype ) with a frequency of about 0 . 01% in Caucasian populations [3] causes blood AAT levels below 30% of normal . This genotype is clearly associated with elevated chronic obstructive pulmonary disease ( COPD ) risk accounting for 1–2% of all cases [4] , [5] . There is also strong evidence that accelerated lung function decline and increased obstructive disease risk can be caused by compound heterozygosity of PI Z and PI S ( PI SZ genotype ) . The case is less clear for PI MZ , PI MS or PI SS genotypes ( PI M standing for the normal allele ) , which cause a less pronounced reduction in AAT concentration , as previous studies produced inconsistent evidence [6]–[9] . Further of note , large-scale genome-wide association studies ( GWAS ) on COPD or on cross-sectional or longitudinal lung function have not identified the SERPINA1 gene to be a major genetic determinant [10]–[12] . But a recent GWAS on emphysema [13] and a comprehensive evaluation of candidate regions for lung function [14] reported rs4905179 and rs3748312 , two common SNPs ( minor allele frequencies ( MAFs ) >5% ) located in the SERPINA gene cluster on 14q32 . 13 , among their most strongly associated results . This locus encompasses SERPINA1 and ten other genes ( SERPINA2 to SERPINA6 and SERPINA9 to SERPINA13 ) encoding extracellular ‘clade A’ serpins with very heterogeneous functions [15] . It is currently not known whether such association signals observed for this locus reflect a causal role of common variants or whether they are merely synthetic , reflecting effects of rarer causal variants [16] . Towards that aim , but also to detect further chromosomal loci of potential relevance to circulating levels of AAT , we first performed a GWAS on AAT serum level using a subset of the population-based Swiss Cohort Study of Air Pollution and Lung Disease in Adults ( SAPALDIA ) as discovery sample , and a second subset of SAPALDIA as well as an independent cohort , the Copenhagen City Heart Study ( henceforth referred to as Copenhagen ) , as replication sample . We also conducted fine mapping analyses of the SERPINA1 gene in the SAPALDIA cohort . Finally , we meta-analyzed the lung function effect of common and low-frequent SERPINA1 SNPs previously observed to be associated with pulmonary health in ever-smokers , based on data provided by several population- and patient-based studies .
The association of more than 2 . 1 million genome-wide SNPs with AAT serum levels is shown in Figure 1 . The ten most strongly associated SNPs were all located in the SERPINA gene cluster , half of them reached genome-wide significance ( P<5*10−8 , Table 1 ) . The top 100 ranking SNPs are provided in Table S3 . A regional association plot for the SERPINA gene cluster is shown in Figure 2 . Both the top-ranking imputed SNP , rs2736887 , and the top-ranking genotyped SNP , rs4905179 , were located in close proximity to the SERPINA6 gene and approximately 33 kb and 50 kb downstream of SERPINA1 ( effect estimates β = −0 . 071 and −0 . 068 g/L per minor allele; P = 2 . 48*10−13 and 1 . 20*10−12 , respectively ) . Linkage disequilibrium ( LD ) between these two variants based on HapMap2 CEU ( Utah residents with Northern and Western European ancestry ) derived haplotype data [18] was strong ( r2 = 0 . 88 , D′ = 1 ) , but Figure 2 suggests that the LD , expressed in r2 values , between the top-ranking SNP and the other SNPs in the region is generally modest . The genomic inflation factor lambda was low ( λ = 1 . 02 ) , suggesting minimal population stratification . The quantile-quantile plot ( Q-Q plot ) showed good adherence to null expectation and substantial positive deviation between observed and expected p-values for the top-ranking SNPs ( Figure S2 ) . In a sensitivity analysis adjusting for additional covariates , including high sensitivity C-reactive protein ( hs-CRP ) , body mass index ( BMI ) , passive smoking and alcohol intake , the genome-wide association results did not show an increase in the strength of the top-ranking loci , nor did they point to additional loci ( data not shown ) . Even though this GWAS was enriched with asthma patients , GWAS stratification according to asthma status did not show heterogeneity for the top-ranking signals between participants with and without asthma ( data not shown ) . In order to further refine association signals in this region , we imputed additional SNPs on chromosome 14 using haplotype data from the 1000 Genomes Project ( 1000G ) [19] . The 1000G imputation yielded a three times higher number of imputed variants with reasonable quality scores ( imputation-r2>0 . 5 ) compared to HapMap-derived imputed variants . A region defined by 1 Mb up- and downstream of the SERPINA1 gene revealed 24 additional variants that were associated below a local significance level of P<3*10−5 , adjusting for approximately 1800 SNPs covering a region of 2 Mb ( Table S4 ) . Among them , four low-frequent variants and one rare variant showed p-values reaching genome-wide significance level , and interestingly , none of them was in high LD ( r2>0 . 8 ) with any other regional variant tested . The most strongly associated signal came from the PI Z variant , which is well known to be associated with reduced AAT serum levels ( β = −0 . 620 g/L per minor allele , P = 4 . 61*10−43 , MAF = 0 . 84% ) . The other well established causal polymorphism , the PI S variant , was less prominently ranked ( β = −0 . 110 g/L per minor allele , P = 1 . 95*10−6 , MAF = 5 . 70% ) and exhibited an insufficient imputation quality ( imputation-r2 = 0 . 45 ) . Accuracy of the imputed PI S and PI Z results was confirmed by direct genotyping of the samples [20] . The discovery arm revealed 33 PI Z carriers , 111 PI S carriers and two compound heterozygous carriers of PI S and PI Z ( MAF = 1 . 26% for PI Z and 4 . 06% for PI S , respectively ) . No homozygous PI S or PI Z genotypes were detected . To test the influence of these variants on the initially reported GWAS results ( Figure 1 ) , we performed a conditional GWAS by additionally adjusting the regression models for the presence of PI S and PI Z alleles . We observed a drastic change in the association of the SERPINA gene cluster SNPs with AAT serum level ( Figure 3 and Table 2 ) . The strong signal on chromosome 14 observed in the original GWAS disappeared completely and the top-ranking imputed and genotyped SNPs ( rs2736887 and rs4905179 ) were no longer significant ( P = 0 . 44 and 0 . 31 , respectively ) . In fact , no SNP was found near the SERPINA gene cluster among the 100 most strongly associated common variants ( Table S5 ) . In addition , the 1000G imputed data , comprising sequences 1 Mb up- and downstream of the SERPINA1 gene , did not show evidence of other independent AAT-associated SNPs . An alternative approach that excluded all PI S and PI Z carriers from the GWAS sample ( N = 146 ) , instead of adjusting for them , confirmed the results . Both analyses revealed an intergenic region on chromosome 3 with borderline genome-wide significance ( top-ranking SNP rs2566347 , β = −0 . 043 g/L per minor allele , P = 7 . 88*10−8 , in the adjusted GWAS ) . The top SNPs in this region were located in proximity to MFSD1 and RARRES1 , which are two genes with sparsely annotated function . The 1000G imputation of this region did not reveal further variants . In addition , we were unable to replicate this association signal in the SAPALDIA replication arm ( N = 4245 , β = −0 . 004 g/L per minor allele , P = 0 . 46 ) . The effect of rs4905179 , the top genotyped SNP in our GWAS , on AAT serum levels was tested for replication in Copenhagen ( Table 3 ) . The minor allele was associated with β = −0 . 097 g/L ( P<0 . 0001 , N = 8332 ) . As observed in the GWAS , adjustment for PI S and Z polymorphisms resulted in a complete loss of this signal ( β = 0 . 003 g/L , P = 0 . 57 , N = 8273 ) . In a first fine mapping step , 16 SERPINA1 SNPs ( see Materials and Methods section for a description of the SNP selection ) were successfully genotyped in 5569 SAPALDIA subjects ( discovery and replication arm combined ) . The genotype results in the discovery arm allowed us to compare allele frequencies with imputed results derived from the 1000G data . Table S6 shows that the agreement was very high . Stepwise conditional regression analyses were then applied to evaluate the independent effects of each of these SNPs on AAT serum levels ( Table 4 ) . The PI Z variant was most strongly associated with circulating levels of AAT . The PI S variant remained strongly associated after conditioning on PI Z . Two variants located in the 5′ non-coding gene region ( rs2896268 and rs1956707 ) were marginally associated with the phenotype in two further steps after conditioning on PI S and PI Z . The total variance of AAT explained by statistical models increased from 8 . 8% ( model with only non-genetic factors ) to 32 . 6% ( adding PI S and PI Z alleles ) , and to 32 . 8% adding rs2896268 and rs1956707 . Based on genotype data from the SAPALDIA cohort , the SERPINA1 gene contains three haplotype blocks using D′-based block definition ( Figure 4 ) . The AAT deficiency variants PI S and PI Z are located in block 1 , while rs2896268 and rs1956707 are located in block 3 , roughly 8 kb upstream of exon 1 . In a second fine mapping step , exon sequencing was performed in 410 subjects with low AAT levels that were independent of the presence of PI S or PI Z alleles [21] . 16 additional SERPINA1 variants ( two deletions and 14 SNPs ) were detected , of which all but one had already been described [21]–[29] ( Table S7 ) . Three of the SNPs were synonymous , and five had no accession numbers in public databases ( as of April 1st , 2013 ) . Most of the non-synonymous SNPs have already been described as potentially lowering AAT serum level , and computational tools only classified one of them as no damaging to the protein's tertiary structure . In order to estimate the phenotypic influence of these rare variants , we compared mean AAT blood levels , adjusted for sex , age , study center , current smoking , as well as for the presence of PI S and Z alleles , between samples without rare variants ( N = 346 ) and those with a single rare variant ( N = 63 ) or more than one ( N = 1 ) . The subjects with rare variants had a lower adjusted mean AAT level ( 0 . 904 g/L , 95% CI 0 . 884 to 0 . 924 g/L ) compared to those without rare variants ( 0 . 992 g/L , 95% CI 0 . 984 to 1 . 000 g/L , P<0 . 001 ) . Although this difference is small , the range covers the recently proposed upper limit of intermediate AAT deficiency ( 0 . 92 g/L ) , a value with some clinical relevance [20] . AAT levels of carriers of synonymous mutations or non-synonymous mutations without predicted damaging consequences to protein structure ( N = 20 ) were not different from those carrying no rare variants ( 0 . 985 vs . 0 . 990 g/L , P = 0 . 77 ) . Assuming that unsequenced samples were negative for mutations with predicted deleterious functional effects , the total variance of explained AAT further increased from 32 . 8% to 35 . 4% ( based on a statistical model adding all rare mutations with predicted damaging consequences to the protein structure ) . Results from a previous GWAS on emphysema [13] and a large-scale evaluation of candidate loci on lung function [14] pointed to a role of common variants in the SERPINA gene cluster . The SNPs rs4905179 ( associated with emphysema in smokers [13] ) and rs3748312 ( associated with cross-sectional lung function among ever smokers [14] ) were strongly associated with AAT in our study ( Tables 1 and 4 ) , but both signals disappeared upon adjustment for the low-frequent variants PI S and Z . In order to clarify whether the association of the two common SNPs with pulmonary health could also be explained by effects of the rarer SNPs , we conducted a meta-analysis for cross-sectional lung function in ever-smokers across 17 studies with a total sample size of N = 24 , 446 ( Table S2 ) . We included nine studies which had contributed to the original finding on lung function [14] and had available genotypes or 1000G imputed genotype data on PI S and Z . The meta-analysis in cohorts of general population study design showed that rs4905179 was not associated with lung function in ever-smokers ( P = 0 . 90 in the fixed-effect meta-analysis , N = 20 , 153 , Figure 5 ) . Yet smaller studies recruited within population isolates showed a trend for the rare allele to be associated with low lung function ( random-effect P = 0 . 02 , N = 1623 , Figure 5 ) , and in contrast to the association with AAT serum levels , adjusting for PI S and Z alleles did not modify the association of rs4905179 with lung function ( Figure 6 ) . For the second common SNP , rs3748312 , we could nominally replicate the statistically significant allele effect on FEV1 in the general population of ever-smokers ( P = 0 . 02 , N = 15 , 450 ) , and the stronger effect that was published [14] seems to be driven by population isolates ( Figure 7 ) . Again , as for rs4905179 , the associations were not dependent on S and Z alleles ( Figure 8 ) . Meta-analyses of the associations of PI S and Z alleles with lung function revealed no consistent associations between these functional AAT level determining variants and reduced FEV1 ( Figures S3 and S4 ) . Remarkably , the significant associations of rs4905179 and rs3748312 with lung function assessed in two additional studies with patients featuring compromised pulmonary health and undergoing lung resection , showed evidence for synthetic associations of the common SNPs with lung function that are consistent with our results for circulating AAT ( Table 5 ) . The minor alleles were associated with lower lung function and the association completely disappeared when conditioned on the presence of PI S and Z alleles .
We present here the first GWAS on circulating AAT blood levels . Our results confirm that genetic variation in the SERPINA1 gene is a strong determinant of serum AAT levels . Fine mapping of SERPINA1 and subsequent stepwise regression analyses further revealed that the associations with common variants in the SERPINA locus could be attributed to rarer variants previously identified to be causally linked with AAT deficiency . There is an ongoing debate about whether rare variants are responsible for the missing heritability observed in GWAS on many complex outcomes [30] . We show here an example in which the polymorphisms PI S and PI Z seem to account for basically all observable effects of common variants in the SERPINA gene cluster on AAT serum level . The top-ranking genotyped SNP in our GWAS , rs4905179 , was in low r2-based LD with PI S ( r2 = 0 . 18 ) and PI Z ( r2 = 0 . 06 ) , reflecting in part the unequal allele frequencies of these SNPs . However , PI S and PI Z showed very high LD in terms of D′ with the GWAS top signals ( e . g . D′ = 0 . 95 and 0 . 96 , respectively , with rs4905179 ) and generally with many common variants in this locus ( Figure 4 ) , suggesting little genetic recombination . This proof-of-principle approach , revealing that signals of common variants in fact merely reflect rarer variants , has recently also been shown for some of the loci regulating low-density lipoprotein ( LDL ) cholesterol [31] , [32] . Yet for other loci linked to LDL cholesterol , as well as for loci influencing other traits , both common and low-frequent variants contributed independently of the original GWAS signal to the phenotypic trait [31] , [33] , [34] . Using regional 1000G imputation within the top-ranking loci can allow the identification of additional association signals of stronger size to support the initial GWAS top result , as observed here for the SERPINA cluster , but not for the locus near MFSD1 , an association which was not confirmed in the SAPALDIA replication arm . The resequencing strategy of the GWAS-identified locus in a sample with low AAT concentrations yielded in the identification of rare variants being strongly associated with reduced AAT blood levels . Such an accumulation of rare variants in the extreme range of the respective phenotype has also been reported by others [35] , [36] . As for the relative contribution of genetic variants on the phenotype , we confirmed that effect sizes of PI S and Z on AAT serum levels were comparably strong , explaining alone a high proportion of the total variability ( 24 . 2% ) . We estimated that rare variants explained at least another 2% in our population-based sample , but since we did not sequence the entire SAPALDIA sample for rare variants , we cannot reliably quantify this contribution . In terms of blood markers , similar examples exist in which one genetic variant could explain well above 5% of the phenotype's variability ( e . g . lipoprotein ( a ) [37] , bilirubin [38] or adiponectin [39] ) , but for many other markers like serum lipid levels , only variants with small effects have been detected so far [40] . Association patterns between SERPINA1 variants and circulating AAT did not translate to according associations with lung function level in a straightforward manner . Lung function is a complex phenotype associated with numerous genetic variants [11] , [12] . Studies on the associations of SERPINA1 polymorphisms with lung function and COPD have produced mixed results . It is well accepted that severe AAT deficiency caused by PI null mutations or by the presence of two PI Z alleles puts a subgroup of carriers at higher risk of emphysema and COPD , especially when smoking [5] . Studies on COPD found suggestive evidence for an association with heterozygous status for the PI Z allele [6] , [8] , but we observed no associations in our meta-analysis between PI Z and lung function level in ever-smokers . In the SAPALDIA general population sample , we had previously reported that an effect of the PI Z allele on lung function decline is restricted to persistent smokers and primarily observed for forced expiratory flow 25–75% [9] . Other variation in or close to the SERPINA1 gene has been proposed to play a role for pulmonary health . First , a haplotype pattern of five common SNPs was reported to be more frequent in COPD cases than in controls in a study with limited statistical power [41] . The only SNP which was also separately associated with COPD in that analysis was not associated with reduced serum levels in our study ( rs8004738 , Table 4 ) . Second , the minor allele of rs4905179 , which was the top signal in the current AAT GWAS , was positively associated with emphysema assessed by chest tomography in three independent cohorts consisting of smoking COPD patients without severe AAT deficiency ( PI ZZ ) [13] . Finally , the minor allele of the intronic SNP rs3748312 was positively associated with lung function in ever-smokers from different population-based studies of the SpiroMeta Consortium [14] . The association of these two SNPs with lung function in ever-smokers was heterogeneous across studies in our meta-analysis . Dependency on PI S and Z was limited to studies in patients with lung resection ( Groningen , UBC ) , consistent with the notion that SERPINA1 may only confer risk in selected population subgroups . There are several possible explanations for the poor translation of genetic association patterns with serum AAT to lung function and for the heterogeneity of associations between SERPINA1 variants and lung function . First , lung function is influenced by mechanisms in addition to protease-antiprotease disequilibrium . Second , the contribution of the SERPINA1 gene variants as a determinant of lung function likely depends on both the evolutionary pressure in isolated populations and the prevalence of effect modifiers in the respective study populations . These include smoking and smoking intensity , and likely other markers of inflammation . AAT itself plays a dual role in its relationship with lung function . While chronic AAT deficiency is etiologically associated with adverse pulmonary health , individuals with lung function impairment in fact exhibit higher AAT levels for a given genetic background due to AAT's role as an acute-phase inflammation marker [42] , [43] . Third , tissue-specific regulation of the SERPINA1 locus may play an important role . Serum AAT levels are driven by SERPINA1 expression , protein formation and secretion in hepatocytes , so that regulatory SNPs associated with serum AAT likely reflect processes in the liver . One way to infer causality of potentially regulatory SNPs is by testing if they are simultaneously associated with health outcome and gene expression in the relevant tissue [44] . We therefore conducted a look-up in an expression quantitative loci ( eQTL ) database of lung tissue [45] , but could not find any common variant which was significantly associated in cis with the transcripts deriving from the SERPINA1 locus . In a recent study on networks of blood metabolites , the SNPs rs11628917 and rs1884549 were the most strongly associated blood and liver eQTLs with respect to SERPINA1 expression [46] . They both lie in the 3′ untranslated region of SERPINA1 , but were not associated with blood AAT in our GWAS ( P = 0 . 80 and P = 0 . 21 , respectively ) . Moreover , we could not detect epistasis between those variants and the deleterious coding variants PI S and Z in terms of AAT serum levels . The absence of such an interaction does not point to regulatory function of the common SNPs [47] and argues in favor of tissue-specific heterogeneity . Forth , the role of SERPINA1 in selected subgroups of persons exhibiting accelerated lung function decline or COPD needs to be considered from a perspective beyond genetic variation , as a recent study investigating epigenetic mechanisms of disease revealed methylation status of the SERPINA1 gene to be most strongly associated with cross-sectional lung function and COPD [48] . The strength of this study is that it combines the report of a GWAS on AAT serum levels with meta-analyses of the associations of some of the GWAS top variants with lung function . The effects of the underlying functional variants are thoroughly investigated resulting in the hitherto largest meta-analysis of PI S and Z on FEV1 in ever-smokers . Ascertainment and study design of the many participating studies were sufficiently diverse to informatively address heterogeneity in association of common and rarer variants in the SERPINA gene cluster with lung function . The strength of the discovery sample is the population-based study design and the detailed characterization of the participants . Sex , age and smoking are important modifiers of AAT blood levels in the general population [43] and were included in all regression models . More refined smoking variables covering smoking intensity were not included as this information is less complete than smoking status in SAPALDIA and would lower the sample size . By excluding samples with elevated hs-CRP values we avoided the masking of AAT deficiencies due to a chronic or acute inflammation . On the methodological side , conditional analysis is a well-established tool for identifying independent signals within a certain locus [38] , [49] , [50] . Furthermore , 1000G imputation was able to point to the causal variant demonstrating its reliability to correctly assign alleles close to the 1% MAF threshold . The limitations of this investigation include firstly the small sample size of the GWAS discovery arm , resulting in a high susceptibility to false negative findings . We calculated 63% power to detect SNPs with an allele effect of 0 . 1 g/L AAT serum level ( = 2 . 4% of the phenotypic variance ) to a genome-wide significance level of 5*10−8 . However , if we define the clinically important threshold of AAT as the upper limit of intermediate AAT deficiency , which has been recently suggested as 0 . 92 g/L [20] , we have more than 99 . 9% power to detect such a large-impact variant . Nevertheless , genes that contribute to AAT serum levels with smaller effects than SERPINA1 were likely to be missed . This could be a reason why neither SNPs in interleukin 6 ( IL-6 ) nor in hepatocyte nuclear factor 1α ( HNF-1α ) /HNF-4 , both important regulators of AAT expression [51] , were associated with circulating AAT concentrations . Furthermore , by sequencing only the coding region of SERPINA1 , rare variants in introns and outside the gene could not be determined . Another potential limitation of our GWAS on AAT serum level is the overrepresentation of asthmatics in the discovery sample . Asthma patients usually show higher levels of inflammatory markers in their lungs . However , we did not find heterogeneity in the effects of the most strongly associated SNPs when comparing asthmatics with non-asthmatics . Moreover , AAT mean values between the discovery and the replication arm were not significantly different , as participants with elevated hs-CRP had been excluded . In conclusion , our study confirms the SERPINA1 locus as the major genetic determinant of AAT blood levels . Methodologically , it represents a powerful example how low-frequent variants , separated by several kilobases from the top-ranking GWAS signals , can create purely synthetic associations which do not add to the variance of the respective outcome . In terms of lung function , our data do not support a functional role of any common SNP in the SERPINA cluster in the general population .
SAPALDIA was approved by the Swiss Academy of Medical Sciences , the national ethics committee for clinical research ( UREK , Project Approval Number 123/00 ) and the Cantonal Ethics Committees for each of the eight examination areas ( Ethics commissions of the cantons Aargau , Basel , Geneva , Grisons , Ticino , Valais , Vaud and Zurich ) . Participants were required to give written consent before any part of the health examination was conducted either globally ( for all health examinations ) or separately for each investigation . For ethics statements of the additional studies contributing to this work , see Table S8 . AAT serum levels in SAPALDIA were determined by latex-enhanced immunoturbidimetric assays ( Roche Diagnostics , on a Roche Cobas Integra analyzer ) with interassay coefficients of variation below 5% and lower detection rate of 0 . 21 g/L . Serum concentrations in Copenhagen were measured by immunoturbidimetric assays ( Thermo Scientific , on a Thermo Scientific Konelab analyzer ) with coefficients of variation below 5% and lower detection rate of 0 . 10 g/L . Lung function was measured in all participating studies by spirometry without bronchodilation ( [52] and Table S8 ) . In the patient-based studies , in which a lung resection was carried out ( Groningen , UBC ) , lung function measurements were carried out prior to the intervention . In an attempt to find AAT modifying SERPINA1 gene variants acting independently of each other , a multiple strategy to optimally cover the gene was applied . Sequencing of the whole SERPINA1 gene in 25 unrelated samples from the Italian registry of AAT deficiency which demonstrated extreme phenotypes was used to identify common SNPs not present in HapMap . Extreme phenotypes consisted of 11 samples with AAT>1 . 60 g/L and hs-CRP <8 mg/L , 3 samples with PI ZZ or PI SZ genotype and AAT<0 . 20 g/L , 2 samples with PI MZ genotype and AAT<0 . 60 g/L , as well as 9 non-carriers of PI S or PI Z alleles with blood levels >0 . 65 and <1 . 10 g/L . In these 25 samples , a total of 129 mutations were identified in the SERPINA1 gene . After removing SNPs which were monomorphic in our data , SNPs deviating from HWE or lying in high LD with an adjacent marker ( D′>0 . 8 and r2>0 . 4 according to JLIN [57] ) , we finally obtained a list of 22 common SNPs ( Table 4 , selection A ) . In a second strategy , HapMap CEU data was used to select tagging SNPs ( Haploview 3 . 32 ) [58] , resulting in 8 polymorphisms ( selection B ) . Third , TAMAL [59] was used to identify promising SNPs in the region of the SERPINA1 gene ( selection C ) . Pairwise LD and the feasibility of designing a corresponding TaqMan assay reduced the number of SNPs to 13 . Two established ( PI S and Z ) and one suggestive ( rs8004738 [60] ) functional SNPs were added ( selection D ) , resulting in 16 SNPs used in the conditional analysis . Three of them were already part of the SNP array genotyped for the GWAS . The 5 SNPs in coding regions ( exons 2–5 ) were all non-synonymous . We sequenced 410 individuals with abnormally low AAT levels with the Sanger chain-termination method . Different thresholds according to the deficiency genotypes and hs-CRP values were applied to define an abnormally low AAT concentration ( PI MM: 1 . 13 g/L if hs-CRP >8 mg/L and 1 . 00 g/L if hs-CRP ≤8 mg/L; PI MS: 0 . 85 g/L; PI MZ: 0 . 65 g/L ) [21] . The cut-off of 1 . 13 g/L was earlier reported to be the best to differentiate AAT-deficient patients from healthy individuals [61] . Since exon 1 is non-coding , the sequencing procedure was only applied to exons 2 to 5 . AAT serum levels were only marginally skewed to the right , and a log-transformation of these data was omitted since it led to a stronger deviation from normality . Student's t-test was used to compare adjusted mean AAT levels between different subgroups of the sequenced samples . The genome-wide association of 2 . 17 million quality-controlled SNPs with serum AAT levels was assessed using fixed effects linear regression with ProbABEL [62] . An additive genetic model was applied and the association was adjusted for sex , age , study center , dichotomous current smoking status , as well as population stratification factors . To account for population stratification , we relied on previously inferred ancestry-informative principal components using EIGENSTRAT 2 . 0 software [63] and HapMap data , as well as additional reference European samples [64] . Cryptic relatedness was detected based on identity-by-state ( IBS ) analysis . Influence of additional suggestive determinants of AAT , such as hs-CRP , BMI , alcohol intake and passive smoking was assessed in a sensitivity analysis . We also performed genome-wide analysis conditioned on the functionally established PI S and PI Z variants . Bonferroni correction for multiple testing was applied , resulting in P<5*10−8 to designate genome-wide significance , taking account of one million independent tests for common variants across the genome . For the SNPs imputed by using 1000G reference samples , we considered a three times lower p-value as adequate as roughly three times more SNPs on chromosome 14 passed an imputation-r2 threshold of 0 . 5 ( 219 , 471 1000G-derived variants vs . 82 , 296 HapMap2-derived variants ) . Applying this to a 2 Mb chromosomal stretch ( with approximately 600 HapMap2-derived SNPs ) resulted in a significance threshold of roughly 3*10−5 . For the replicated SNP in the SAPALDIA replication arm , as well as for the lung function analysis , a two-sided p-value of 0 . 05 was considered significant . We investigated heterogeneity between asthmatics and non-asthmatics in the discovery arm by testing for a difference between the two effects , using a chi-square test with one degree of freedom . Replication analysis for AAT in Copenhagen , as well as association analyses of the 16 genotyped SERPINA1 SNPs in both the SAPALDIA discovery and replication arm , was carried out applying the same statistical model as in the GWAS apart from the adjustment for population stratification factors . Stepwise conditional analyses were conducted by testing each SNP for AAT association after including at each step the most significantly associated SNP in the model . As some of these SNPs turned out to be in unexpectedly high LD , we applied a threshold level for statistical significance of P = 0 . 005 , accounting for approximately ten independent tests [65] . To be as close as possible to the calculations carried out in the original publication [14] , multivariate linear regression models for lung function analyses were used adjusted for sex , age , height and population stratification factors ( if available ) . All the SAPALDIA regression analyses were performed with STATA 12 . 1 IC . Manhattan , Q-Q and forest plots were created with the help of R 2 . 15 . 1 ( www . r-project . org ) . Regional association plots were drawn using LocusZoom [66] . Pairwise LD was calculated for HapMap2 and 1000G CEU data using SNAP [67] . The LD plot was produced with HaploView 4 . 2 [58] . The effect of non-synonymous SNPs on protein structure was predicted by SIFT [68] . Finally , Quanto 1 . 2 . 4 ( hydra . usc . edu/gxe/ ) was used for power calculations for the GWAS .
|
Low levels of alpha1-antitrypsin ( AAT ) in the blood are a well-established risk factor for accelerated loss in lung function and chronic obstructive pulmonary disease . While a few infrequent genetic polymorphisms are known to influence the serum levels of this enzyme , the role of common genetic variants has not been examined so far . The present genome-wide scan for associated variants in approximately 1400 Swiss inhabitants revealed a chromosomal locus containing the functionally established variants of AAT deficiency and variants previously associated with lung function and emphysema . We used dense genotyping of this genetic region in more than 5500 individuals and subsequent conditional analyses to unravel which of these associated variants contribute independently to the phenotype's variability . All associations of common variants could be attributed to the rarer functionally established variants , a result which was then replicated in an independent population-based Danish cohort . Hence , this locus represents a textbook example of how a large part of a trait's heritability can be hidden in infrequent genetic polymorphisms . The attempt to transfer these results to lung function furthermore suggests that effects of common variants in this genetic region in ever-smokers may also be explained by rarer variants , but only in individuals with hampered pulmonary health .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genome-wide",
"association",
"studies",
"medicine",
"epidemiology",
"genetics",
"biology",
"human",
"genetics",
"genetics",
"of",
"disease",
"pulmonology",
"genetic",
"epidemiology"
] |
2013
|
Causal and Synthetic Associations of Variants in the SERPINA Gene Cluster with Alpha1-antitrypsin Serum Levels
|
Leishmania parasites , the causative agent of leishmaniasis , are transmitted through the bite of an infected sand fly . Leishmania parasites present two basic forms known as promastigote and amastigote which , respectively , parasitizes the vector and the mammalian hosts . Infection of the vertebrate host is dependent on the development , in the vector , of metacyclic promastigotes , however , little is known about the factors that trigger metacyclogenesis in Leishmania parasites . It has been generally stated that “stressful conditions” will lead to development of metacyclic forms , and with the exception of a few studies no detailed analysis of the molecular nature of the stress factor has been performed . Here we show that presence/absence of nucleosides , especially adenosine , controls metacyclogenesis both in vitro and in vivo . We found that addition of an adenosine-receptor antagonist to in vitro cultures of Leishmania amazonensis significantly increases metacyclogenesis , an effect that can be reversed by the presence of specific purine nucleosides or nucleobases . Furthermore , our results show that proliferation and metacyclogenesis are independently regulated and that addition of adenosine to culture medium is sufficient to recover proliferative characteristics for purified metacyclic promastigotes . More importantly , we show that metacyclogenesis was inhibited in sand flies infected with Leishmania infantum chagasi that were fed a mixture of sucrose and adenosine . Our results fill a gap in the life cycle of Leishmania parasites by demonstrating how metacyclogenesis , a key point in the propagation of the parasite to the mammalian host , can be controlled by the presence of specific purines .
Protozoan parasites from Leishmania genus are the causative agents of leishmaniasis , a broad spectrum disease that range from asymptomatic infections to disfiguring forms such as diffuse or mucosal leishmaniasis as well as visceral leishmaniasis , which can be fatal if not adequately treated . The outcome of this infection in humans depends largely on the immune response assembled by the host and the virulence and species of the parasite . Basically developmental stages of these protozoa alternate between amastigotes that live in mammalian macrophages and generate the disease manifestations mentioned above , and promastigotes which are present in the midguts of female sand flies . During its life cycle in the invertebrate host , Leishmania promastigotes undergo a series of morphological changes that culminate with the differentiation into the metacyclic form , which is responsible for the initiation of infection in the vertebrate host . Although this developmental stage has been described for nearly 30 years [1] , the factors that trigger metacyclogenesis in Leishmania parasites are still poorly understood . It has been generally stated that “stressful conditions” will lead to development of metacyclic forms and with the exception of a few studies no detailed analysis of the molecular nature of the stress factor has been performed [2] . Based mainly in in vitro studies , it has been demonstrated that low pH , lack of nutrients and low levels of tetrahydrobiopterin influence metacyclogenesis [1] , [3] , [4] . However , no specific role of these factors in vivo has ever been confirmed . Leishmania and other trypanosomatids are unable to synthesize the purine ring by the de novo pathway and depend on the uptake of nucleosides and nucleobases to supply the purine salvage pathways [5] . The present study reports that metacyclogenesis induction is controlled by the presence of adenosine . We observed that addition of CGS 15943 ( CGS ) , a potent antagonist of mammalian adenosine receptors [6] , strongly induces metacyclogenesis in promastigote cultures . We also show that although CGS interferes with the transport of adenosine by the parasite , induction of metacyclogenesis cannot be attributed to lack of precursors for the purine salvage pathway and does not correlate with lack of parasite proliferation . Furthermore , we show that addition of adenosine to cultures of metacyclic promastigotes induces differentiation of these cells into proliferative stages of the parasite . Finally , we show that the presence of adenosine in the sugar meal of infected sand flies inhibit metacyclogenesis indicating that the effect of adenosine on metacyclogenesis is not restricted to the development of the parasite in vitro .
The protocols to which animals were submitted were approved by the Universidade Federal de Ouro Preto ( OFíCIO CEP N° . 005/2009 ) and by the Universidade Federal de Minas Gerais Ethical Committees on Animal Experimentation ( 125/05 and 211/07 ) and followed the guidelines from the Canadian Council on Animal Care . Leishmania amazonensis [PH8 strain ( IFLA/BR/67/PH8 ) ] and Leishmania major [FRIEDLIN strain ( MHOM/IL/80/Friedlin ) ] were cultured in Grace's insect medium ( Sigma Aldrich ) supplemented with 10% heat-inactivated fetal calf serum ( FCS; LGC Biotecnologia ) , 2 mM L-glutamine ( GIBCO BRL ) and 100 U/ml penicillin G potassium ( USB Corporation ) , pH 6 . 5 , at 25°C . Parasites were sub-cultured at 1×105 parasites/ml 3 days before experiments in order to achieve a mid-log phase . Throughout this work , promastigotes cultures were initiated with 2×106 cells/ml and incubated for 3 days . Parasites cultures were initiated from frozen stocks ( first passage ) obtained after the transformation of C57BL/6 mice lesion amastigotes . In vitro cultures of differentiated promastigotes , for all experiments , were maintained in conditions described above for a maximum of ten passages . This limit of passages is adopted to prevent loss of metacyclic characteristics , such as virulence , as previous described [7] . Viability of parasites was assessed by flow cytometry using propidium iodide incorporation [8] . Parasites were washed twice , resuspended at 2×106/ml in PBS and , at the moment of acquisition , 5 µl/ml of a propidium iodide staining solution ( eBioscience ) was added to samples . Data were collected in BD FACSCalibur flow cytometer . Cell acquisition was performed using BD CellQuest Pro software and data analysis was performed using FlowJo software ver . 9 . 4 ( Tree Star , Inc ) . Fifth thousand events were harvested from each sample . Female BALB/c and C57BL/6 mice ( 4–8 weeks old ) were obtained from the University's animal facility ( CCA - UFOP ) . A mongrel dog of unknown age naturally infected with Leishmania infantum chagasi from an endemic area of visceral leishmaniasis in Minas Gerais State ( southeast Brazil ) was housed in a kennel at Universidade Federal de Minas Gerais animal facility . Animals were given water and food ad libitum . Adult females of Lutzomyia longipalpis were maintained in a closed colony in UFMG as described [9] . Unless otherwise stated in figure legends , CGS 15943 ( Tocris Bioscience/Sigma-Aldrich ) ( 50 µM ) was added to culture after 48 hr of growth and metacyclic promastigotes quantified after 24 hr . In experiments of metacyclogenesis induction , nucleosides [adenosine ( 0 . 5 mM ) , inosine ( 0 . 5 mM ) ] , nucleobases [adenine ( 0 . 5 mM ) and hypoxanthine ( 0 . 1 mM ) ] , dipyridamole ( 0 . 05 or 0 . 1 mM ) and N6-methyladenine ( 2 mM ) were also added alone or with CGS ( see figure legends ) after 48 hr of culture and incubated for 24 hr . CGS 15943 , Dipyridamole , N6-Methyladenine and hypoxanthine were prepared in DMSO ( LGC Biotecnologia; 1–3% final concentration ) . Morphological evaluation of metacyclogenesis was performed under light microscopy . Parasites were considered metacyclic when presenting small body cell size and long flagellum ( twice or more the body length ) . Metacyclic promastigotes were enriched by centrifugation of promastigotes over Ficoll 400 ( Sigma-Aldrich ) gradient and quantified by hemocytometer counting [10] , [11] . Results reflect the total number of parasites obtained from the gradient . In general the percentage of metacyclic promastigotes in the enriched fraction was greater than 80% . Promastigotes obtained from control or CGS-treated cultures were incubated in Hank's balanced salt solution prepared with 1 mM MgCl2 and 0 . 15 mM CaCl2 in the presence of 10% of fresh rat serum . After growing parasites as mentioned above , cells were washed twice and resuspended at 1×108/ml in HBSS , pH 7 . 4 plus 10% rat serum and incubated at 37°C for 1 h . Reaction was stopped diluting the samples 100-fold with ice-cold HBSS and centrifuge cells at 1540× g/10 min/4°C . Parasite survival was assessed by counting whole cells on a hemocytometer . These procedures were adapted from [12] . Total protein extracts were prepared [13] and samples equivalent to 2×107 promastigotes were fractionated in 15% SDS-PAGE and transferred to nitrocellulose membranes . Immunoblotting was performed as described [13] with anti-META1 polyclonal antiserum diluted 1∶500 . Membranes were developed using chemiluminescent substrate ( SuperSignal , Thermo Scientific ) . Promastigotes were washed twice with 0 . 9% NaCl solution and adjusted to 2×107 cells/ml . Intestines of adult females of Lutzomyia longipalpis , maintained as described [9] , were dissected ( 10 per group ) , and posterior portion and Malpighi tubules were removed . After this , midguts were opened and placed into 30 µL 0 . 9% NaCl solution plus 1% hemoglobin in a scooped glass slide ( Sigma Aldrich ) . Hemoglobin 0 . 5% final concentration was used to prevent the parasite attachment to the glass . 30 µL of L . amazonensis suspension were added to midguts and incubated for 35 min/25°C . After washing twice with saline , midguts were homogenized with a teflon pestle in micro-centrifuge conical tubes and fixed on slides for Giemsa staining . Midgut homogenates were evaluated for total number of promastigotes by optical microscopy . These procedures were adapted from [14] . Thioglycolate elicited peritoneal cells were harvested and plated ( 1×106 cells ) onto round cover-slips in supplemented DMEM in 24-well plates and rested overnight at 37°C , 5% CO2 . Non-adherent cells were removed by washing with warm PBS . Parasites were added for 3 hr ( 5 parasites per macrophage ) , washed and incubated for another 72 hr at 37°C/5% CO2 . After 72 hr of infection , coverslips were fixed in methanol for 2 min ( Vetec Fine Chemistry ) , and stained using the kit Panótico Rápido ( Renilab ) - a Romanowsky like stain , according to manufacturer's instructions for the assessment of cellular parasitism . The analysis was performed using an Olympus BX50 optical microscope . The number of infected and uninfected cells was determined in a minimum of 100 macrophages per coverslip . C57BL/6 mice were inoculated in the left ear with 1×103 promastigotes/10 µL PBS originated from control or CGS-treated cultures . Lesion development was followed weekly with a digital caliper ( Starrett , model 727 ) . The results were expressed as the difference between measures of infected and contralateral non-infected ears . The number of parasites in the ear was estimated by a limiting dilution assay [11] , [15] . After seven weeks of infection , mice were euthanized and the ears collected , the ventral and dorsal dermal sheets separated and incubated , dermal side down , in RPMI-1640 medium , pH 7 . 2 ( Sigma-Aldrich ) with collagenase A ( 1 mg/mL ) ( Sigma-Aldrich ) for 2 hr at 37°C/5% CO2 . Ears were ground in Grace's insect medium , pH 6 . 5 , in a glass tissue grinder . Tissue debris was removed by centrifugation at 50× g/4°C/1 min and supernatant transferred to another tube and centrifuged at 1540× g/4°C/15 min . The pellet was resuspended in 0 . 5 ml Grace's insect medium supplemented with 10% heat-inactivated FCS , 2 mM L-glutamine and 100 U/ml penicillin G potassium , pH 6 . 5 . Parasite suspension was serially diluted in 10-fold dilutions in duplicates to a final volume of 200 µl in 96-well plates . Pipette tips were replaced for each dilution . Plates were incubated for 15 days at 25°C and examined under an inverted microscope for the presence of parasites . Results were expressed as log of the last dilution in which they were detected . Uptake of [3H]adenosine by L . amazonensis promastigotes was assayed as [16] , [17] with few modifications . Promastigotes from middle log phase cultures were washed twice in buffer ( 116 mM NaCl , 5 . 4 mM KCl , 5 . 5 mM glucose , Hepes/Tris 30 mM , pH 7 . 4 ) and resuspended at 5×108 cells/ml in the same buffer . Parasites were incubated in for 20 min with or without CGS 15943 or dipyridamole ( 10 , 50 or 100 µM ) . Transport was measured at 25°C and initiated by adding 100 µL of cells to 100 µL of radiolabeled adenosine at 2 µM/0 . 2 µCi , diluted in buffer containing CGS 15943 or dipyridamole at the concentrations described above . After 60 s , transport was stopped by spinning the cells ( 10000× g/120 s ) through an oil cushion of 100 µl of dibutyl phthalate ( Sigma-Aldrich ) . Aqueous and oil phase were removed and pellet dissolved with 2% Triton X-100 ( Sigma-Aldrich ) . Lysate was mixed with 2 ml of scintillation solution ( Optiphase HiSafe 3 - PerkinElmer ) for liquid scintillation counting ( Tri-Carb 2810 TR - PerkinElmer ) . To evaluate metacyclogenesis in vivo , adult insects of Lutzomyia longipalpis , majority of females , were allowed to feed in a naturally Leishmania infantum chagasi-infected dog for 30 min . Insects were fed for 8 to10 days with 30% sucrose solution with or without adenosine ( 5 mM ) . Midguts were dissected , homogenized individually with a teflon pestle in micro-centrifuge conical tubes and fixed on slides for Giemsa staining . Each midgut was evaluated for the presence of metacyclic promastigotes by microscopy . Student's t-test was performed using Prism 5 . 0 software ( GraphPad Software ) . p<0 . 05 was considered statistically significant .
Acession number of META-1 protein is available from GenBank ( http://www . ncbi . nlm . nih . gov/genbank ) as AAC04758 . 1 .
|
Leishmania parasites are the causative agent of a spectrum of diseases characterized by severe lesions in skin or life threatening visceral infections . In the parasite life cycle , a range of morphological transitions can be found such as amastigotes ( hosted in humans and others mammals ) and promastigotes ( located in the sand fly vector ) . The disease begins when the infective non-dividing promastigote metacyclic form is transmitted to mammalian hosts by the bite of an infected sand fly . The conditions for the development of these metacyclic promastigotes are still not fully understood . Here we tested the hypothesis that the presence or absence of purine will determine whether the parasite will proliferate or differentiate into the metacyclic form . Our experiments indicate that the presence of purines in the culture medium of Leishmania parasites interfere with the development of metacyclic promastigotes . Our results also show that although decreased proliferation and metacyclic differentiation occur simultaneously these two phenomena are independently regulated . Finally we show , for the first time in a natural vector , that metacyclogenesis is inhibited in vivo by adenosine , suggesting a new level of cell differentiation control in these protozoan parasites which is driven by purine sensing .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"Discussion"
] |
[
"parastic",
"protozoans",
"leishmania",
"parasitology",
"protozoology",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"microbial",
"growth",
"and",
"development"
] |
2012
|
Leishmania Metacyclogenesis Is Promoted in the Absence of Purines
|
Gene loci are found in nuclear subcompartments that are related to their expression status . For instance , silent genes are often localized to heterochromatin and the nuclear periphery , whereas active genes tend to be found in the nuclear center . Evidence also suggests that chromosomes may be specifically positioned within the nucleus; however , the nature of this organization and how it is achieved are not yet fully understood . To examine whether gene regulation is related to a discernible pattern of genomic organization , we analyzed the linear arrangement of co-regulated genes along chromosomes and determined the organization of chromosomes during the differentiation of a hematopoietic progenitor to erythroid and neutrophil cell types . Our analysis reveals that there is a significant tendency for co-regulated genes to be proximal , which is related to the association of homologous chromosomes and the spatial juxtaposition of lineage-specific gene domains . We suggest that proximity in the form of chromosomal gene distribution and homolog association may be the basis for organizing the genome for coordinate gene regulation during cellular differentiation .
The nucleus appears to be organized according to the many functions it performs [1 , 2] . The nucleolus , for example , is a subcompartment that exists as a result of its activities: rDNA transcription and ribosomal biogenesis [1] . Gene loci reflect this functional organization in that their subnuclear localization often correlates with their expression status . Among many examples , it has been demonstrated that: ( 1 ) silent loci positioned at the nuclear periphery relocalize to the nuclear center when activated during cellular differentiation ( e . g . , [3 , 4] ) ; ( 2 ) subsets of expressed genes from a single chromosome territory ( CT ) colocalize in transcription factories [5]; and ( 3 ) the regulation of cell-type–specific genes correlates with their association in the nucleus , despite being found on different chromosomes [6] . In addition , gene loci are often localized relative to their respective CT , with active gene domains looped away from the territory and inactive domains at its surface ( e . g . , [7 , 8] ) . These observations and others have rekindled interest in a long-standing question in the study of nuclear organization: do chromosomes have defined positions within the nucleus ? Structural arrangements of chromosomes , such as the Rabl configuration and the prometaphase rosette , have been known for some time , and there are recent examples of the nonrandom organization of chromosomes [9] . Although it has become clear that nuclear organization is inherently probabilistic , the tendencies for certain chromosomes to be preferentially localized within the nucleus have been demonstrated . For example , analysis of the radial positioning of individual CTs within human nuclei revealed that gene-dense chromosomes have a propensity to be centrally localized , whereas gene-poor chromosomes are more peripheral [10–12] . This phenomenon has also been observed in the nuclei from other primates [13] . An examination of the organization of all chromosomes within individual human nuclei , however , did not reveal a consistent role for gene density in CT localization [14] . Rather , this analysis determined that a chromosome's size ( as a function of its overall length ) is also related to its radial positioning , with small chromosomes being found more centrally positioned . Similar results were observed in an analysis of mouse nuclei [15] . The varying impact of chromosome density and size may be due to cell-type differences or to the method of analysis ( e . g . , focusing on a chromosome's center of gravity as opposed to its total area or volume ) . Nevertheless , a common basis for nonrandom chromosome organization beyond basic chromosome characteristics such as gene density or overall length has yet to be elucidated . Analysis of genomes from multiple species has revealed that genes have a particular linear arrangement along chromosomes: the co-regulated genes of transcriptomes have a marked tendency to be found grouped ( or clustered ) according to their shared expression status [2 , 16–18] . Therefore , gene loci not only localize to positions within the nucleus relevant to their expression , but they are also inherently organized nonrandomly along chromosomes . It is unclear what function this clustering of genes plays , although the prevailing model suggests that clusters create expression “hubs” or “neighborhoods” in which the linearly proximal genes alter the dynamics of regulatory protein ( transcription factor ) binding by increasing the relative abundance of binding sites [19] . Given that hundreds of genes are regulated during cellular differentiation , the localization of individual gene clusters may be reflected in the organization of chromosomes enriched in these co-regulated genes [20] . Specifically , chromosomes may be organized in relation to their total , cell-specific expression profile . This organization may involve nuclear localization of chromosome territories , interchromosomal interactions , or both . We have used an in vitro model of murine hematopoiesis to test the hypothesis that cellular differentiation is associated with a relationship between the linear arrangement of co-regulated genes and chromosomal organization . FDCPmix cells are nontransformed , multipotential hematopoietic progenitors that can be maintained and differentiated into a number of blood cell lineages—including highly pure populations ( ∼90% ) of erythrocytes and neutrophils—with the appropriate cytokines [21] . To explore the possibility of a link between gene expression and genomic organization , we determined the linear chromosomal arrangement of co-regulated genes and the organization of chromosomes for the three cell types . Our results demonstrate a relationship between the chromosomal distribution of co-regulated genes and the propensity for homologous chromosomes and co-regulated gene domains to be proximal . We suggest that the spatial proximity of genes along chromosomes and the association of homologous chromosomes help ensure the coordinate regulation of genes during cellular differentiation .
Using data from a microarray analysis of gene expression along a time course of differentiation ( Figure S1A ) [22] , we analyzed the linear chromosomal distribution of co-regulated genes from the FDCPmix cells ( hereafter referred to as progenitors ) and the derived erythroid and neutrophil cell types . By using Affymetrix databases , the National Center for Biotechnology Information ( NCBI ) mouse genome alignment ( 32v1 ) , and BLAT ( BLAST-like alignment tool ) run locally , we assigned 93% of the ∼12 , 000 genes represented on the MG-U74Av2 chip to their linear chromosomal positions . We next assigned the erythroid and neutrophil co-regulated genes ( or gene sets ) to their linear positions . To determine whether the clustering of co-regulated genes is also true in a mammalian differentiation model , we compared the linear distribution of the 594 erythroid and 539 neutrophil genes with lineage-specific expression patterns ( Table 1 and Figure S1B ) to a simulated gene set , created by the random positioning of “genes” in the ∼11 , 000 assigned microarray positions and iterated 1 , 000 times ( Materials and Methods ) . Performing a χ2 analysis with the simulated gene set and those of the lineages , we observed that the frequency of co-regulated genes grouped without an interceding unregulated microarray gene is significantly larger in each lineage than predicted by the simulation ( p < 0 . 0001 ) ( Figure 1A ) . We found that 18% ( 106/594 ) of erythroid and 20% ( 106/539 ) of neutrophil genes are found in clusters of two and—to a lesser degree—three . Interestingly , the examples from other species of co-regulated gene clustering found similar percentages [2] . We verified that the clusters were not due to duplications by removing all redundant GenBank accessions and eliminating any microarray sequences which overlapped with more than one gene ( using BLAT ) or had any shared sequence identity . These results show that the spatial proximity of co-regulated genes extends to vertebrates and the differentiation of multipotent progenitors extends to derived cell types . Furthermore , the majority ( 77% ) of the co-regulated genes in the erythroid lineage are down-regulated ( Figure S1B ) , demonstrating that in addition to gene activation , clustering may play a role in gene silencing . Considering the expanse of the entire genome , gene tandems and triplets represent relatively small stretches of DNA . To examine the linear organization of genes beyond clusters , we performed a sliding-window analysis to compare the entire erythroid , neutrophil , and simulated genomic gene distributions to that of the microarray ( Figure 1B ) . Our sliding-window approach—in which a 10–megabase pair ( Mbp ) window is moved in 1-Mbp increments—helps overcome the relative infrequency of lineage-specific/simulated genes compared with the gene number represented on the microarray chip ( ∼600 versus ∼11 , 000 ) . Furthermore , the 10-Mbp window provides a biologically relevant frame , because comparing the murine and human genomes revealed the two share syntenic domains of ∼10–15 Mbp , which suggests a functional constraint on gene domain size [23] . In a proportions test , the distribution of erythroid and neutrophil co-regulated gene sets differed significantly from the microarray ( p < 1 . 2 × 10−35 ) , whereas the simulated gene set did not ( p < 0 . 22 ) ( Figure 1B ) . The tendency of the lineage gene sets toward gene-dense domains drives their difference with the simulation ( Figure 1B , insert ) . Therefore , beyond tandems and triplets , there is an inherent propensity for the lineage-co-regulated genes to exhibit genomic proximity in domains . To visualize the gene distributions , we plotted the erythroid , neutrophil , and microarray sliding-window data along the chromosomes ( Figure 2; for all chromosomes , see Figure S2 ) . As the Chromosome 7 example indicates , the mouse chromosomes display regions that are gene dense and gene poor both for the microarray and the co-regulated gene sets from the two lineages ( Figure 2 ) . This chromosome structure was first characterized in the analysis of the human genome , wherein gene-dense domains were shown to also be regions of increased gene activity ( RIDGEs ) , whereas the gene-poor regions ( or valleys ) have little gene activity [24] . In our comparison , there are a number of lineage-specific regions with significantly greater and fewer co-regulated genes than expected by the microarray ( Figure S2 ) . For example , the region between ∼ 61 and 71 Mbp is enriched in both lineages for genes that share lineage-specific regulation—either silencing ( erythroid ) or activation ( neutrophil ) ( Figure S3 ) . Despite these significant differences , however , the gene distributions of the lineages generally follow that of the overall microarray profile , because the significant difference in gene distributions exhibited in the proportions test described above imply a greater tendency for co-regulated gene density , not that the domains are necessarily different from the total gene distribution ( depicted on the microarray ) . Regardless , these data indicate that large-scale , nonrandom gene domains characterize the linear structure of murine chromosomes as well as the distribution of co-regulated genes during differentiation . There are many examples of gene loci demonstrating activity-dependent nuclear localization; therefore , we hypothesized that the complement of expressed genes in the three cell types of our differentiation model may exhibit an inherent nuclear localization pattern . To test this idea , we generated probes for fluorescence in situ hybridization ( FISH ) that detect total gene expression in the progenitor , erythroid , and neutrophil cell types ( Figure 3A ) ( Materials and Methods ) . We prepared double-stranded DNA from cDNA prepared from each lineage and used it to amplify probe material through a modified protocol for chromatin immunoprecipitation ( ChIP ) microarray analysis [25] , incorporating either biotin- or digoxigenin-conjugated nucleotides . We analyzed the percentage of probe material represented in three concentric nuclear shells of equal area in two-dimensional ( 2-D ) images ( Text S1 ) . In all three cell types , the hybridizations revealed the preferential localization of active genes in the inner nuclear shells , with the innermost shell making up the majority of probe signal ( Figure 3B ) . Importantly , the probe materials that were produced with two different conjugated nucleotide tags were concurrently detected in each nucleus to verify the hybridization patterns ( Figure 3A ) . Therefore , beyond the examples of individual loci , lineage-specific gene expression appears to be spatially organized in the nuclear center . These results are in agreement with bromodeoxyuridine ( BrdU ) -incorporation analysis , which has indicated that early-replicating chromatin ( active or euchromatic ) is centralized in the nucleus , whereas late-replicating chromatin ( silent or heterochromatic ) is enriched in the nuclear periphery ( e . g . , [26] ) . Furthermore , the concentration of active gene expression may parallel the role of proximity in the linear chromosome organization of co-regulated genes described above . As indicated in the introduction , previous studies of human chromosomes have alternately found density or length playing a role in their radial localization within the nucleus . In contrast , our analysis above indicates a preference for expressed genes to localize to the nuclear center . Although mouse chromosomes are more uniform than their human counterparts , they still vary widely in their degree of density and length ( Table 1 ) . To evaluate the relative importance of these characteristics , we determined the localization of CTs in our differentiation model by performing 2-D FISH on the three lineages with a representative battery of whole-chromosome probes ( or paints ) ( 2 , 3 , 4 , 5 , 6 , 7 , 11 , 12 , 14 , 17 , and 19 , which include short , long , and gene-dense/-poor chromosomes ) ( Table 1 ) . As in the above analysis of gene expression , we measured the percentage of CTs in three concentric nuclear shells of equal area ( Figure 4A ) . Unexpectedly , all chromosomes showed a significant enrichment in the central portion of the nucleus when compared with the middle or outer shells , regardless of cell type or chromosome size/density ( analysis of variance ( ANOVA ) , p < 0 . 0001 ) ( Figure 4B; for individual chromosomes , see Figure S4 ) . The inner and middle regions together compose the vast majority of each CT area . We corroborated our results in nuclei prepared to preserve their 3-D structure ( Figure S5A ) , and we found no significant difference between the 2-D analysis and the CT localization in concentric shells of the six faces of the nuclei sphere ( Figure S5B–S5D ) . That the outer region demonstrates that the lowest percentage of CT area may be linked to the observation that the nuclear periphery is enriched in heterochromatin [27] , which is not detected by chromosome paints . Moreover , the central localization of chromosomes may be related to this region's described transcriptional permissiveness ( Figure 3A ) [2] . The presence of nucleolar organizing regions does not account for this localization , because in the mouse , rDNA is found primarily on the smallest chromosomes and we do not see a relationship with size [27] . Although all the chromosomes we analyzed have their area enriched in the nuclear center , three of the five densest mouse chromosomes ( 2 , 7 , and 17 ) demonstrate an even greater concentration in the inner region ( Figure S4 ) . Since these chromosomes are gene dense , they also have a proportionately large number of the lineage-specific genes ( Table 1 and Dataset S1 ) . Therefore , the demonstrated tendency of gene-dense human chromosomes to be localized in the nuclear center may be due to their having the greatest number of active genes in any particular cell type . Further research will be necessary to understand fully the radial organization of CTs and the function it may play in coordinate gene regulation . When analyzing CT radial distribution , we observed the tendency for homologous chromosomes to be in proximity of each other ( Figure 4A ) . Therefore , using the images from the radial analysis , we measured the interaction of homologs through intensity thresholding , with CTs being scored as associated only if the above background pixels were unambiguously connected ( Figure 4C ) . This stringent criterion reveals that chromosomes show a high frequency of homologous interaction in the interphase nucleus . An average of 50% of nuclei in each lineage display homologous chromosome association , with variations among individual chromosomes ( Figure 4D; for individual chromosomes , see Figure S6A ) . Furthermore , the results are not due to the 2-D approach , because we also analyzed homolog association through the depth of nuclei prepared to preserve their 3-D structure and found no significant difference in the results ( Figure S7A and S7B ) . The association for pairs of heterologous chromosomes was also measured , demonstrating a high degree of interaction ( on average ∼40% ) ( Figure 4D; for each pair , see Figure S6B ) . However , because there is twice the possibility of interaction between two heterologous chromosome pairs than a single homologous pair , these data support the prevalence of homologous chromosome association . We suggest that homolog proximity may be related to the propensity for CTs to be localized to the nuclear center and to the chromosomal distribution of co-regulated genes . To determine whether the association of homologs is correlated to their number of co-regulated genes , chromosome gene density , or chromosome size ( length ) , we performed a Kruskal-Wallis ( K-W ) test comparing these chromosomal attributes ( Figure 4D and Table 1 ) . A multivariate statistical analysis , the K-W test is a one-way ANOVA by ranks , in which each dataset is ranked in a column—according to chromosome number—and then statistically analyzed in rows across values ( or the conditions of co-regulated gene number , proximity , and length ) . For the three lineages , there is a striking pattern of significance in that the proximity of homologous chromosomes is related only to the chromosomal distribution of co-regulated genes ( Table 2; for the rankings , see Figure S8 ) . The erythroid cells , e . g . , reveal that the ranking of chromosomes for homologous proximity does not significantly differ from the ranking of chromosomes for their distribution of co-regulated genes , yet it does for the basic characteristics of size and density ( Table 1 ) . Therefore , the distribution of co-regulated genes appears uniquely related to the proximity of homologous chromosomes , underscoring the importance of proximity in coordinate gene regulation . Coordinate gene expression has been thought of as a type of network , because it is composed of genes ( or nodes ) that are related ( linked ) in terms of their co-regulation and shared function , such as in the differentiation of a given cell type . In real-world networks—shown to be prevalent in biological systems—a diminishing number of nodes with an increasingly greater number of links create a hub organization [28] . The lineage-specific erythroid and neutrophil linear chromosomal gene distributions exhibit this characteristic , with their sliding window data demonstrating a negative power-law degree-distribution ( erythroid P ( k ) = k–2 . 6 , neutrophil P ( k ) = k–2 . 5 ) . This behavior provides a basis for modeling gene regulation during differentiation , emphasizing the importance of linear proximity and suggesting that spatial proximity of gene domains may also play an important role in coordinate gene regulation . Specifically , the prevalence of homolog association may be related to the proximity of the co-regulated gene clusters within the nucleus . To test this hypothesis , we arbitrarily identified five gene domains with unique coordinate gene expression in the erythroid and neutrophil cell types on two chromosomes of relatively equal length ( 2 and 4 ) ( Figure 5A ) . We analyzed the spatial proximity of the homologous domains—determined as a ratio of distance between domains to nuclear diameter—in the progenitor , erythroid , and neutrophil lineages using 2-D FISH ( Figure 5B; verified in 3-D in unpublished data ) . The co-regulated genes in these domains are all active in the progenitors . Consistent with this shared expression status , the domains do not demonstrate significant differences in their separation in the progenitor nuclei ( ANOVA p = 0 . 49 ) ( Figure 5C ) . Importantly , however , the spatial proximity of the domains in the erythroid and neutrophil nuclei do differ significantly ( ANOVA , p < 0 . 001 and p < 0 . 01 , respectively ) ( Figure 5C ) . In both lineages , the degree of domain proximity varies according to its overall activity status . For example , the domains without any regulated genes demonstrate the greatest separation , whereas the most active domains are the closest . Analyzing the data for the degree of direct loci colocalization supports the overall behavior of the domains ( Dataset S1 ) . Therefore , these results support the hypothesis that the spatial proximity of lineage-specific gene domains may further facilitate the co-regulation of genes colinear along chromosomes . Given the complexity of chromosome distribution in the interphase nucleus , prior attempts to determine the simultaneous organization of all chromosomes have relied on center-of-gravity measurements [14] . However , this type of analysis does not take into account the contours of a CT , which are relevant in discerning chromosome associations . Therefore , we developed a strategy to analyze the simultaneous relationship of all chromosomes in prometaphase rosettes , when a cell's complement of chromosomes come together to form a circle with their centromeres ( Figure 6A ) . We used spectral karyotyping ( SKY ) [29]—developed for the clinical detection of chromosomal abnormalities—and implemented a method to perform pattern recognition on SKY rosettes in SVCell software ( SVision , Bellevue , WA , United States ) . Our approach executes distance-constrained , zone-of-interest ( ZOI ) region partition on the SKY image ( Figure 6B ) , from which chromosome proximity can be automatically determined for all chromosome associations at a resolution of one pixel ( for a complete description of the software , see Text S1 ) . Rosettes have long been used to study chromosome organization ( e . g . , ( 14 , 30] ) , although it remains controversial whether chromosome relationships are maintained through mitosis [31–33] . However , regardless of whether organization is maintained , we examined rosettes to determine if the tendency for homolog proximity is observed under conditions that permit analysis of all chromosomes at the same time . Since the mouse genome is composed of two complements of 19 autosomes and two sex chromosomes , the likelihood that a chromosome associates with any other in the rosette is at least 2/39 ( or 5 . 1% , associations can occur on either side of the chromosome ) ( Figure 6A ) . To examine a region of chromosome association , we defined proximity as being no more than two chromosomes apart along the contours of the territories ( Figure 6A ) . Using this criterion , we compared the association of every chromosome to all others in simulated ( Materials and Methods ) and lineage rosette datasets . In support of our findings from individual chromosome analysis ( Figure 4C and 4D ) , we observed a high frequency of proximity for homologous chromosomes in the three cell types: on average , homologs associated in 48% , 51% , and 40% of progenitor , erythroid , and neutrophil rosettes , respectively ( Figure 6C; for individual chromosome data , see Figure S9A ) . In comparison , only 11% of homologs associated in the simulated rosettes ( ANOVA , p ≤ 3 . 8 × 10−11 ) , which reflects the random expectation for our designation of proximity ( 6/39 or 15 . 4% ) ( Figure 6C ) . Importantly , a significant difference is maintained whether proximity is defined as chromosomes being directly adjacent or one chromosome removed ( Dataset S1 ) . Homologous chromosomes vary in their degree of proximity among the three cell types , although there is no size-dependent trend ( Figure S9A ) . An earlier study of rosettes , using individual chromosome paints , had found that homologs tend to be located across the center of the rosette ( or transversely related ) [34] . To exclude this possibility and to verify our observation of proximal association , we measured the number of homologs separated by at least 2 . 618 radians ( or an angle of 60° ) across the rosette center from the chromosome analyzed ( Figure 6A ) . The simulated rosette set closely follows the prediction of 21% for random association ( a 60° angle includes ∼8 chromosomes , 8/39 , 24% ) ( Figure 6C ) . Homologous chromosomes in the lineages , however , show a significantly lower degree of transverse separation than the simulated dataset does ( ANOVA , p < 4 . 7 × 10−8 ) , corroborating our determination of proximity ( Figure 6C; for individual chromosome data , see Figure S9A ) . A K-W test comparing proximal homologs , transverse homologs , chromosome gene density , and chromosome size support the conclusion that co-regulated gene distribution is uniquely related to the association of homologs ( Table S1 ) .
By combining analyses of gene expression patterns and chromosome localization , we have tested the hypothesis that coordinate gene regulation during cellular differentiation is related to a specific organization of the genome . Like examples from other organisms , we found that genes co-regulated during murine hematopoiesis are significantly colinear , forming gene clusters along chromosomes ( Figure 1 ) . Furthermore , we determined that clustering is not limited to gene activation , because the erythroid lineage is characterized by gene silencing and displays a similar degree of clustering as neutrophils do . Beyond the adjacency of individual genes , we found a wide-spread tendency for the co-regulated gene sets to reside nonrandomly in large gene domains ( Figure 2 ) . Reasoning that the examples of individual loci exhibiting lineage-specific nuclear positions may be broadly reflected in active gene localization during cellular differentiation , we developed an approach to determine the nuclear distribution of a cell type's complement of expressed genes . Our results revealed that the nuclear interior is not only transcriptionally permissive , but the preferred region for coordinate gene expression ( Figure 3 ) . This pattern of localization is mirrored in the chromosomes themselves , with CTs being enriched in the nuclear center ( Figure 4B ) . Interestingly , this preference for central positioning of CTs is coupled with a propensity for homologous chromosomes to interact ( Figure 4D ) . The degree of homolog association is related to the chromosomal distribution of co-regulated genes ( Table 2 ) , and representative gene domains analyzed by FISH exhibit greater spatial proximity in the nucleus according to their lineage-specific expression patterns ( Figure 5 ) . Finally , by using a novel means of analyzing the simultaneous organization of chromosomes , we corroborated the tendency for homologs to be proximal ( Figure 6 ) . Therefore , despite its complexity and probabilistic nature , the nucleus appears to be nonrandomly organized for coordinate gene regulation . Our analysis suggests that the co-regulated gene distributions of the erythroid and neutrophil lineages can be described as scale-free networks . Beyond their temporal regulation and chromosomal distributions , the lineage-specific gene domains also demonstrate physical proximity within the nucleus , underscoring their regulatory linkage . An important feature of networks—in particular those that are scale-free—is their tendency for self-organization . We suggest that transcriptional regulation during cellular differentiation is related to the nucleus self-organizing according to the principle of proximity [1] . Extending studies demonstrating that X inactivation is related to the physical interaction of the X chromosomes [35 , 36] , we argue that the association of homologous chromosomes is widespread and correlated with the proximity of similarly regulated gene domains during cellular differentiation . Therefore , homolog association may facilitate the formation of expression hubs containing the alleles of co-regulated gene domains [2] . Hematopoietic progenitors have been shown to undergo “lineage priming , ” a low-level promiscuous expression of genes expressed in differentiated cell types [37 , 38] . The role that lineage priming may play in gene expression during differentiation remains to be determined . However , given the high diffusion rate of regulatory [39] and structural [40] nuclear proteins , it is attractive to consider that spatial proximity would alter the off-rate for DNA-binding proteins , creating localized protein concentrations—such as in the nuclear center—to ensure the co-regulation of relevant gene sets . Therefore , beyond its central role in the homologous recombination that helps fuel variation and natural selection , diploidy may also be involved in facilitating the co-regulation of entire gene sets during cellular differentiation . Whether allelic proximity is a requirement for or a result of transcriptional regulation and the mechanism ( s ) underlying the association of homologs remain to be established .
For the culture and analysis of FDCPmix cells , the progenitors were routinely cultured in Iscove's Modified Dulbecco's ( GIBCO ) medium supplemented with 20% ( vol/vol ) horse serum ( GIBCO ) and 10 ng/ml recombinant murine IL-3 ( R&D Systems ) . Differentiation of the progenitors was performed as previously described [21] . The growth factor concentrations used were as follows: erythroid: Epo ( 5 U/ml; Amgen ) , hemin ( 0 . 2 mM; Sigma ) , and rmIL-3 ( 0 . 05 ng/ml; R&D ) ; neutrophil: G-CSF ( rmG-CSF; 50ng/ml; R&D ) and SCF ( rmSCF; 100 ng/ml; R&D ) . Cells were stained with benzidine and cytospins stained with May-Grünwald-Giemsa to verify cellular morphology . In addition , cell lineages were verified on a FACS-Vantage ( Becton Dickinson ) after staining for cell surface markers , with fluorescein-conjugated mouse monoclonal antibodies directed against Ter-119 , Gr-1 , or Mac-1 ( Pharmingen ) . Combining information from Affymetrix databases and the NCBI mouse genome alignment ( 32v1 ) with BLAT run locally , we assigned 93% of the ∼11 , 000 genes ( 11 , 160 ) represented on the MG-U74Av2 chip to their linear chromosomal positions . Checking for GenBank duplicate entries , we also assigned lineage ( erythroid or neutrophil ) and differential expression classes ( I–IV , zero indicates no change ) to these positions ( Dataset S1 , ComboDatabase sheet ) . The simulated dataset was created by randomly positioning 650 “genes” in the ∼11 , 000 microarray positions , iterated 1 , 000 times . The number 650 was originally chosen , because this was the larger gene set size ( of the erythroid and neutrophil ) before removal of duplicate accession on the microarray chip . Using data from this combined dataset , we performed statistical analysis using the R statistical environment or the statistical package XLSTAT ( Addinsoft ) . Data processing— i . e . , sliding window , data simulation , or position assignment—was performed using a combination of perl scripts and R . Statistical analysis was performed by extracting relevant data from this larger dataset . The K-means , proportions test , and χ2 analyses were performed directly . When a test required an intermediate output from the perl scripts , these intermediate files ( e . g . , sliding-window datasets ) are in Dataset S1 . Intermediate files were used to demonstrate power law distributions and to perform exact binomial tests ( Dataset S1 , Chr#_E&N sheets ) . Using Prism ( Graphpad Software ) we performed a K-W test with a Bonferroni post-test on the nonparametric rosette data ( Dataset S1 , RosetteData sheet ) and ANOVA with Bonferroni post-test for the masked territories ( Dataset S1 , Mask_Regions%Territory sheet ) , among other analyses . Progenitor , erythroid , and neutrophil cells were taken from asynchronously growing cultures at day five of the differentiation . Slides were prepared by either briefly treating with a hypotonic solution ( KCl , 0 . 075 M ) and then fixing in at least five changes of 3:1 absolute methanol:glacial acetic acid before spreading , or first adhering the suspension cells to chambered slides using poly-L-lysine followed by hypotonic treatment and fixation . 2-D FISH was performed as described [3] with slight modifications . Briefly , slides were prepared for DNA hybridization by treatment with RNase ( 100 mg/ml ) , ethanol washes , and subsequent denaturation in 50% Formamide/50% 2xSSC at 75 °C for 2 min . To analyze genomic gene expression within the nucleus , we first isolated total RNA from the progenitor , erythroid , and neutrophil cells using TRIzol Reagent ( Invitrogen ) . Subsequently , we generated double-stranded ( ds ) cDNA using SuperScript ( Invitrogen ) following the manufacturer's protocol . Importantly , the number of cycles for the amplification has to be empirically determined for each sample . We found that 17 cycles provided a defined spread of products of sufficient quantity . Following ds cDNA production , we slightly modified the protocol used for the amplification of probe material in chromatin immunoprecipitation ( ChIP ) -microarray analysis [25] , by using ds cDNA as input and incorporating biotin- and digoxigenin-conjugated nucleotides through random labeling . Whole chromosome probes ( or paints ) to various chromosomes ( 2 , 3 , 4 , 5 , 6 , 7 , 11 , 12 , 14 , 17 , and 19 ) were obtained from either Applied Spectral Imaging or Cambio and were hybridized according to manufacturer's specifications . Images of single or multiple Z sections were captured on an Olympus IX 70 or a Zeiss Axiovert 100 TV microscope equipped with cooled CCD cameras and operated using Deltavision SoftWorx software ( Applied Precision ) . Deconvolution was performed using AutoDeblur , or in some cases , no deconvolution was used on the images ( AutoQuant Imaging ) . Masking occurred using Imaris software ( Bitplane ) . Masks were created by intensity thresholding to remove background and capture the extent of the gene expression probe material or CT area . For each fluorescent channel , these masks were converted to binary masks and then either exported as TIFs for nuclear region analysis using SVCell ( SVision ) ( Text S1 ) or assayed for homologous and heterologous CT association ( identified by unambiguous contiguity of CTs ) . At least 30 nuclei were assayed for each type of analysis . For a detailed description of the 3-D FISH protocol , see Text S1 . Image stacks of 30–40 Z sections ( spaced 0 . 25 μm apart ) were captured on a Zeiss Axiovert 100 TV microscope equipped with a cooled CCD camera and subsequently deconvolved using AutoDeblur . The 3-D images were analyzed as a projection of the six faces of the nucleus ( CT localization ) or as a volume ( CT association ) in Imaris . Random real-world networks have been shown to follow a negative power-law degree-distribution , P ( k ) ∼ k–g , with a g ( degree exponent ) between 2 and 3 . We plotted on a log-scale the number of links ( k ) , which in our analysis are the 10-Mb domains from the sliding-window analysis , as a log function of the frequency of those domains with a given density . The distributions of the gene sets conform to the expectations of a scale-free network , with degree exponents of 2 . 6 ( erythroid ) and 2 . 5 ( neutrophil ) ( Dataset S1 ) . Analysis of 1- and 5-Mb domains yields similar distributions ( unpublished data ) . Rosettes were enriched from asynchronous populations by the preparation of slides the day after splitting the cultures . To be analyzed , the rosettes had to exhibit the characteristic circular shape formed by the centromeres without pronounced perturbations; at least 30 rosettes per lineage were analyzed . The simulated rosettes were created from ten rosettes from the various lineages with their karyotype information removed . Using a random number generator ( http://www . randomizer . org ) , we made 100 sets of random numbers ( a set consists of two random lists of 1–20 ) , and moving from one chromosome to another imposed the random number as its chromosome karyotype . SKY hybridization and detection were performed according to manufacturer's specifications ( Applied Spectral Imaging ) . SVCell alpha prototype software ( SVision LLC ) was used to assay spatial pattern associations between individual chromosomes across all the rosette images from the three lineages and a simulated rosette image set . SVCell is a microscopy image informatics tool that contains fast image recognition algorithms , relational measurements , and supports the creation and review in real time of a large number of spatial patterns that can be derived from these relational measurements . SVCell image recognition and relational measurements automatically normalize the distortion and intersample variations among input images . We created spatial pattern rules in SVCell and used them to interrogate the rosette images for the pattern's frequency in total and across all chromosome interactions . The application is described in detail in Text S1 .
|
How are genomes—and the chromosomes that comprise them—organized in the eukaryotic nucleus ? This long-standing question in cell biology has gained renewed interest due to observations that gene regulation is correlated with the nonrandom distribution of gene loci linearly along chromosomes and spatially within the nucleus . We have used an in vitro model of cellular differentiation to test the hypothesis that there is an inherent organization of the genome related to coordinate gene regulation . Our analysis reveals that during the differentiation of a murine hematopoietic ( blood-forming cell ) progenitor to derived cell types , co-regulated genes have a marked tendency to be proximal along chromosomes in the form of clusters ( of two and three genes ) and large-scale domains . Overall gene expression is also spatially proximal , with a pronounced concentration in the nuclear center . The chromosomes themselves parallel this organization of gene activity , with chromosome territories localizing primarily in the interior of the nucleus . Surprisingly , we found that homologous chromosomes have a tendency to be associated , the extent of which is related to the number of co-regulated genes residing on the particular chromosome . Furthermore , individual gene domains display lineage-specific proximity according to their co-regulation . Our study supports the idea that the eukaryotic nucleus is broadly organized—with proximity playing a key role—to facilitate coordinated gene regulation during cellular differentiation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"cell",
"biology",
"molecular",
"biology",
"hematology"
] |
2007
|
Coordinate Gene Regulation during Hematopoiesis Is Related to Genomic Organization
|
Listeria monocytogenes ( Lm ) infection induces rapid and robust activation of host natural killer ( NK ) cells . Here we define a region of the abundantly secreted Lm endopeptidase , p60 , that potently but indirectly stimulates NK cell activation in vitro and in vivo . Lm expression of p60 resulted in increased IFNγ production by naïve NK cells co-cultured with treated dendritic cells ( DCs ) . Moreover , recombinant p60 protein stimulated activation of naive NK cells when co-cultured with TLR or cytokine primed DCs in the absence of Lm . Intact p60 protein weakly digested bacterial peptidoglycan ( PGN ) , but neither muropeptide recognition by RIP2 nor the catalytic activity of p60 was required for NK cell activation . Rather , the immune stimulating activity mapped to an N-terminal region of p60 , termed L1S . Treatment of DCs with a recombinant L1S polypeptide stimulated them to activate naïve NK cells in a cell culture model . Further , L1S treatment activated NK cells in vivo and increased host resistance to infection with Francisella tularensis live vaccine strain ( LVS ) . These studies demonstrate an immune stimulating function for a bacterial LysM domain-containing polypeptide and suggest that recombinant versions of L1S or other p60 derivatives can be used to promote NK cell activation in therapeutic contexts .
NK cells are lymphocytes that help control infections and tumors and regulate autoimmune responses through both cytotoxic and cytokine-secreting effector functions . Through the use of Fas , TRAIL , and secretion of perforin and granzyme B , NK cells induce apoptosis or lysis of distressed somatic cells . NK cells can also activate or suppress other aspects of an immune response through lysis of antigen-presenting cells or production of immune regulatory cytokines such as IFNγ and IL-10 [1] , [2] , [3] . NK cells are activated by secreted cytokines and by proteins on the surface of target cells or other immune cells . NK cells recognize “self” class I major histocompatibility complex ( MHCI ) cell surface molecules by inhibitory receptors that prevent NK cell activation [4] . Reduced expression of MHCI by stressed or infected target cells can thus relieve this inhibition and lead to NK cell activation . Alternatively , NK cells can be directly activated by recognition of stress-induced ligands through a variety of activating NK cell receptors [5] . In addition to direct recognition of target cells , the activity of NK cells is regulated by cytokines released from macrophages and dendritic cells , such as TNFα , IFNγ , IL-15 , IL-12 , and IL-18 [6] , [7] . IL-18 is a potent stimulus for IFNγ production by ( NK ) cells [8] , [9] Full NK cell activation often requires direct contact of the NK cell with accessory cells , such as dendritic cells ( DCs ) . In some cases , contact of NK cells and DCs permits signaling of co-stimulatory molecules [6] , [10] , [11] , [12] . Contact may also promote efficient transmission of IL-18 and/or IL-12 [13] , [14] , and trans-presentation of IL-15 to “prime” the NK cell [15] , [16] . The balance of both inhibitory and activating signals ultimately determines the extent of NK cell activation and possibly the nature of NK cell effector functions . Systemic infection by numerous bacterial pathogens elicits potent NK cell activation and IFNγ production , but the mechanisms of NK cell activation during bacterial infections are incompletely understood . Infection by Listeria monocytogenes ( Lm ) rapidly activates a large population of NK cells to produce IFNγ [17] , [18] . Lm is a facultative intracellular pathogen of humans and animals [19] . A number of secreted Lm virulence factors that contribute to pathogenicity . One of the two most abundantly secreted Lm proteins is a bacterial hemolysin ( Hly ) called listeriolysin O . Hly is essential for bacterial access to the cytosol of host cells and thus for intracellular bacterial growth and virulence during systemic infection of mice [19] , [20] . The second most heavily secreted Lm protein is called p60 . Expression of p60 also contributes to Lm virulence during systemic infections [17] , [21] , [22] . However , the virulence-promoting function of p60 has been enigmatic . The p60 sequence contains a C-terminal NLPC/p60 domain , two N-terminal LysM domains , and a single N-terminal SH3-like domain . Some NLPC/p60 domains have been associated with endopeptidase activity [23] , [24] , while LysM and bacterial SH3 domains generally bind glycans or proteins [25] , [26] , [27] , [28] . Consistent with autolytic endopeptidase activity , semi-purified p60 protein digested Micrococcus luteus cells [29] , [30] , and crude Lm PGN [17] . We previously hypothesized that Lm expression of p60 might thus contribute to Lm pathogenicity by altering the production of immune modulating muropeptides [21] . Subsequently , an immune modulatory function was associated with Lm expression of p60 . Namely , systemic infections by wt Lm promoted significantly increased NK cell activation when compared to infections by p60-deficient ( Δp60 ) Lm [17] . Here , we confirm that p60 deficiency correlates with impaired NK cell activation in a recently developed cell culture assay system . Furthermore , using recombinant p60 protein and p60-derived polypeptides , we show that p60 protein can indirectly enhance NK cell activation in the absence of additional Lm factors . Purified p60 protein binds to DCs and induces IL-18 secretion , which is required for NK cell activation by p60 in co-culture . The ability of p60 to stimulate DCs for NK cell activation mapped to the first LysM and SH3 domains ( L1S ) of the p60 protein . The L1S region was also sufficient to promote activation of NK cells in vivo when given to naïve mice . In vivo treatment with p60 increased serum IFNγ and reduced susceptibility of recipient mice to infection by the heterologous NK cell-sensitive bacterial pathogen , Francisella tularensis . These data demonstrate that p60 protein boosts NK cell activation during Lm infection through appropriate stimulation of accessory cells and suggest that L1S may be useful to therapeutically manipulate immune responses .
Systemic infections with Δp60 Lm strains elicit weak IFNγ production by NK cells [17] . Likewise , bone marrow dendritic cells ( BMDCs ) infected with Δp60 Lm elicited significantly less IFNγ from co-cultured naïve splenic lymphocytes ( Figure 1A ) . Intracellular staining revealed that NK1 . 1+ cells were responsible for nearly all IFNγ production in these cultures ( Figure 1B-E ) . Multiple independently generated p60 deletion mutants showed a similarly poor ability to induce IFNγ production in these co-cultures ( Figure 1F ) . This weak IFNγ production was restored to wt levels when expression of p60 was restored in the Lm Δp60 mutant using an integrated vector coding for His tagged p60 protein ( Figure 1G ) . The complemented Δp60+p60 strain secreted p60 at levels similar to wild-type Lm based on immunoblotting of culture supernatants ( not shown ) . Reduced NK cell activation in response to Δp60 Lm infection might conceivably reflect reduced bacterial burdens within the infected BMDCs . However , microscopy and cfu plating revealed that the growth rate was identical for wt and Δp60 Lm , as was the percent of infected BMDCs over the course of infection ( not shown ) . Finally , the ratio of cytosolic ( actin-associated ) versus phagosome localized Lm was also similar for the two strains ( not shown ) . Thus , expression of p60 was not required for the invasion or cytosolic replication of Lm in BMDCs , but nonetheless increased the activation of neighboring NK cells . We expressed and purified recombinant His-tagged p60 protein from E . coli using nickel affinity and cation exchange columns . When added to co-cultures of BMDCs and nylon wool non-adherent cells ( NWNA ) prepared from naïve mouse spleens , the purified protein induced IFNγ production ( Figure 2A ) . The recombinant p60 protein was associated with ∼1 ng of E . coli LPS per 1 µg of protein . However , this amount of LPS was insufficient to stimulate IFNγ production when added to the co-cultures without p60 protein ( Figure 2A ) . Moreover , production of IFNγ was not seen in response to treatments with BSA or a His-tagged phage autolysin ( HPL511 ) that was purified from E . coli using a similar procedure and also contained ∼1 ng LPS per µg protein ( Figure 2A ) . To further exclude possible artifacts due to LPS , polymyxin B columns were used to remove LPS from the purified p60 protein . The detoxified p60 was initially insufficient to activate IFNγ production ( Figure 2B ) , suggesting that activation by p60 required priming or maturation of the BMDCs . To test this , BMDCs were pretreated with TLR agonists for three hours before addition of p60 . Pre-stimulation of co-cultures with LPS , the non-toxic LPS analog monophosphoryl LipidA ( MPA ) , or poly I∶C ( PIC ) each sufficed to elicit IFNγ production following p60 stimulation ( Figure 2B ) . None of the priming agents tested stimulated IFNγ production on their own . Based on flow cytometry using intracellular IFNγ staining , NK cells were the major source of IFNγ produced in the co-cultures with primed and p60-stimulated BMDCs ( Figure 2C-E ) . To test whether these NK cells responded directly to the stimulated BMDC , NWNA splenocytes were stained and flow sorted to obtain 97–98% pure populations of NK1 . 1+CD3- NK cells , CD3+NK1 . 1- T cells , and “other” cells ( negative for both NK1 . 1 and CD3 ) . Each sorted population was added to BMDCs ( >90% CD11c+ ) that had previously been treated with LPS and a p60-derived peptide ( peptide described further below ) . As previously shown for Lm-infected co-cultures [8] , the purified NK cells produce IFNγ when cultured alone with stimulated BMDCs ( Figure S1 ) . The amount of IFNγ was not significantly affected by adding back either or both other cell populations present in NWNA splenocyte preparations ( Figure S1 ) . Although T cells did not impact IFNγ production by the NK cells , we observed small amounts of IFNγ production when purified splenic T cells were cultured alone with the stimulated BMDCs ( Figure S1 ) . This likely reflects the ability of memory CD8+ T cells to respond to IL-12 and IL-18 in the cultures [31] ( see below for further discussion of cytokines present in the cultures ) . We conclude that the LPS and p60-stimulated BMDCs were sufficient to activate NK cells in these in co-cultures , and that the other cells present in the NWNA population did not significantly modulate this activation . Stimulation of BMDC with LPS and other TLR stimuli elicits production of cytokines that stimulate DC and NK cells . Detoxified p60 , failed to stimulate IFNγ production by the co-cultures in the absence of priming agents and also failed to induce significant levels of IL-12p70 secretion by BMDC . However , the priming agents PIC and MPA both elicited strong IL-12 production in the co-cultures containing NK cells and BMDC ( Figure 2F ) . In some cases , but not universally , this IL-12p70 secretion was further enhanced by p60 stimulation . Recombinant IL-12p70 , IFNβ , and TNFα each sufficed to prime the production IFNγ by detoxified p60 protein in the absence of TLR agonists ( Figure 2G ) . IL-12 was by far the most potent priming agent , most likely due both to BMDC priming and the enhancement of IFNγ transcription in NK cells [32] . These findings suggested that cytokines produced in response to TLR agonists mediate priming or maturation of the BMDCs , which can then respond to recombinant p60 protein or mediate activation of naïve NK cells in NWNA splenocytes in response to this protein . We next asked how p60 might mediate NK cell activation in co-culture by examining the role of accessory DCs . Addition of p60 protein did not stimulate IFNγ production in the absence of NK cells or when added to NWNA cells in the absence of BMDCs ( [17] and Figure 3A ) . This result suggested two possibilities . Either p60 protein might act on DCs to induce the ability of DCs to activate NK cells , or the protein might be presented to NK cells by DCs for NK cell activation . To investigate whether p60 protein bound to BMDCs , the cells were treated or not with p60 protein , fluorescent beads , or p60 plus beads . After washing , the treated and untreated BMDCs were stained using anti-p60 rabbit polyclonal antisera and a secondary Cy3-labeled anti-rabbit Fab ( Figure 3B-E ) . A punctuate staining pattern was seen on the stained p60-treated BMDCs ( Figure 3C and E ) . Identical results were obtained using two independent anti-p60 polyclonal antibodies ( data not shown ) . This punctuate staining was not observed on stained untreated cells or cells treated with beads alone ( Figure 3B and D ) , nor on sorted NK cells , T cells , or other NWNA splenocytes ( not shown ) . The punctuate staining for p60 did not require detergent permeation of the BMDC membrane ( not shown ) , nor did p60 puncta co-localize with phagocytosed FITC-labeled latex beads ( Figure 3E ) . These data suggest that p60 protein binds to an unknown receptor/s present at or near the surface of BMDCs . We previously reported that contact between DCs and NK cells was required for NK cell activation during Lm-infection [8] . Similarly , contact between the DC and NWNA splenoctyes was required for p60-induced NK cell activation ( Figure 3F ) . It was conceivable that binding of p60 to the DC surface might permit presentation of this protein to NK cells . However , nickel beads coated with a His-tagged p60 were not able to stimulate NWNA cells in the absence of BMDC ( Figure 3G ) . Together , these data suggested that p60 primarily stimulates NK cell activation indirectly , due to its effect on DCs . NK cells might respond to altered MHC I expression and/or upregulation of stress ligands by BMDCs treated with p60 protein [5] , [33] . Thus , we stained BMDCs that had been primed with LPS plus or minus an active p60-derived peptide ( described further below ) and assessed their expression of activation markers ( MHCII ) and several known ligands for NK cell surface receptors ( Figure S2 ) . MHCII expression increased after protein treatment , consistent activation of the BMDC . No down regulation of MHC I was observed and the expression of NKG2D ligands RAE1γ , RAE1δ , and MULT1 were unchanged . There was no change in staining levels for the SLAM family members 1 , 2 , 3 , and 6 . SLAMF 5 staining was slightly reduced after protein treatment , which is likely due to DC activation . These data suggested that NK cell activation by p60 was due to effects of p60 on DCs that were independent of altering expression of these known ligands for NK cell activating and inhibitory receptors . Both cell contact and inflammatory cytokines such as IL-12 and IL-18 modulate NK cell activation and IFNγ production [34] . IL-12 production by BMDC infected with wildtype versus Δp60 Lm was not significantly different ( data not shown ) . Since IL-18 production is essential for NK cell activation by Lm infected BMDCs [8] , we asked whether bacterial expression of p60 effected IL-18 production in infected BMDCs . We found that secretion of IL-18 was significantly reduced in the supernatants of C57BL/6 BMDCs infected with Δp60 Lm ( Figure 4A ) . Consistent with this observation , detoxified p60 protein in combination with PIC strongly simulated IL-18 secretion from BMDCs ( Figure 4B ) . We next evaluated the effects of IL-18 production on IFNγ production in cultures of infected BMDC and NWNA splenocytes . In response to Lm infection , IL-18-/- BMDCs stimulated very little IFNγ production ( Figure 4C ) . Moreover , the amount of residual IFNγ produced in these co-cultures was no longer affected by bacterial expression of p60 . Further , IL-18 expression in BMDCs was additionally required to elicit IFNγ production in co-cultures primed with PIC or MPA and stimulated with detoxified p60 protein ( Figure 4D ) . Together , these data suggest that binding of p60 to BMDC elicits IL-18 secretion , which is required for activation of NWNA splenocytes . The p60 protein has been shown to weakly digest peptidoglycan ( PGN ) [21] , [29] , hence , we previously hypothesized that PGN cleavage by p60 might release muramyl di-peptide ( MDP ) or other bioactive muropeptides [21] , [29] . MDP is detected by NOD2 , which signals through the RIP2 kinase [35] , [36] , [37] , [38] . To test whether MDP generation by p60 might stimulate NK cell activation , we compared the ability of Lm infected B6 and B6 . RIP2-/- BMDC to activate NK cells from B6 mice . Bacterial expression of p60 enhanced IFNγ production in NWNA splenocytes co-cultured with RIP2-/- BMDCs to the same extent as C57B6 BMDCs ( Figure S3A ) . Additionally , purified recombinant p60 stimulated BMDC and NK cell enriched splenocytes co-cultures in the absence of added Listeria PGN . Therefore , generation and detection of the MDP PGN fragment was not required for NK cell activation nor for the ability of p60 to enhance such activation . Like the Bacillius subtilis LytF protein , p60 contains a C-terminal NLPC/p60 domain with a putative catalytic triad of two histidines and a single cysteine residue ( Figure 5A ) . In LytF , the cysteine is essential for endopeptidase activity and permits cleavage of the cross-linking peptide chains in peptidoglycan ( PGN ) [24] . However , NLPC/p60 domains have also been associated with other catalytic functions . To formally test whether the enzymatic activity of p60 was required for stimulation of NK cell activation , we engineered and purified a p60 derivative in which the catalytic cysteine residue was mutated to alanine . The resulting p60C389A mutant protein was purified as for wt p60 and tested for digestion of heat-killed Lm and crude Lm PGN substrates using zymography ( Figure S3B and not shown ) . As previously published [21] , [29] , the wt p60 protein cleaved PGN , although this activity was much weaker than that seen with a control phage lysin ( HPL511 ) . The p60C389A was completely inactive in this assay ( Figure S3B ) , confirming that the cysteine residue was required for PGN digestion by p60 . Nonetheless , the purified p60C389A was as efficient as the catalytically active wt protein for stimulating IFNγ production in co-cultures of NWNA splenocytes and BMDCs ( Figure S3C , Figure 5B ) . These data demonstrate that the enzymatic activity of p60 is not required for its ability to promote NK cell activation . Given that enzymatic activity was dispensable for NWNA splenocyte activation by p60 , we asked whether this activation was associated with NLPC/p60 or other domains . The Lm genome contains a homolog of p60 ( Lm0394 ) with both an SH3 domain and a C-terminal NLPC/p60 domain but lacking the N-terminal LysM domains found in p60 . A His-tagged recombinant Lm0394 protein was unable to activate NWNA splenocytes in co-culture ( Figure 5B ) . Thus , the presence of SH3 and NLPC/p60 domains was not sufficient to confer the ability to activate co-cultures . Additional p60 derivatives were engineered and purified , including an N-terminal fragment ( Np60 ) truncated immediately before the TN repeat region and a C-terminal fragment ( Cp60 ) that comprised the TN repeats and NLPC/p60 domain ( Figure 5A ) . These truncated proteins were purified , detoxified , and tested as for full length p60 . Np60 induced IFNγ production in co-cultures pre-stimulated with either PIC or MPA , while Cp60 failed to induce IFNγ ( Figure 5C ) . Further truncation of the N-terminal region mapped the stimulating activity to a fragment containing the LysM1 and SH3 domains , termed L1S ( Figure 5D ) . The results of our experiments with SH3-domain-containing Lm0394 indicate that the LysM1 domain may be responsible for the activity of L1S . However , efforts to purify the LysM1 or SH3 domains alone have thus far been unsuccessful , suggesting that both domains may be required for conformation and stability . Given that the L1S polypeptide was the minimal active component of p60 identified in our studies , we tested whether the LysM1 domain was necessary for p60-induced co-culture activation during Lm infection . We compared Δp60 mutants complemented with p60 constructs that lacked the LysM1 domain or the linker domain ( LD ) between the SH3 and LysM2 domains ( Figure 5A ) . Both complemented strains expressed and secreted the p60 mutant proteins at levels comparable to wildtype Lm based on immunoblotting of precipitated culture supernatants ( not shown ) . The ΔLysM1 complementation mutant induced low IFNγ levels in co-culture similar to Δp60 Lm infection , while the ΔLD complementation mutant induced IFNγ similar to wild type Lm infection ( Figure 5E ) . Thus , the LysM1 domain appears to be largely responsible for p60-mediated activation of BMDC/NWNA splenocyte co-cultures . The regulation of NK cell activation and responses in vivo may differ from their regulation in our cell culture system . We thus asked whether purified , LPS-associated L1S was sufficient to activate NK cells in vivo when administered to mice by intraperitoneal ( i . p . ) injection . LPS was administered to a second group of mice as a negative control . At 24 h after injecting the L1S or LPS , IFNγ production by both splenic and peritoneal infiltrating NK cells was assessed using intracellular cytokine staining . The data showed that LPS treatment failed to stimulate NK cell activation in the absence of L1S polypeptide . However , there were significant increases in the percentage of NK1 . 1+CD3- cells staining positive for IFNγ in both peritoneum ( Figure 6A , 6C ) and spleen ( Figure 6B ) . The activation of splenic NK cells was more modest than seen in the peritoneum , suggesting the NK cell activation largely occurred locally at the site of L1S injection ( Figure 6B ) . The NK cell activation by LPS-associated L1S was dose-dependent ( Figure S4 ) . We additionally observed that the percent granzyme B positive NK1 . 1+CD3- NK cells was increased in the peritoneal cells in response to L1S treatment ( Figure 6D ) . Hence , we measured cytotoxicity from NWNA splenocytes after co-culture with BMDCs stimulated with LPS with or without L1S . Consistent with the increased granzyme B staining in vivo , L1S significantly enhanced the cytolytic activity of NWNA splenocytes against NK cell-sensitive B16F10 melanoma target cells in vitro ( Figure 6E ) . These data confirmed that the p60-derived polypeptide was bioactive in the treated animals and suggested that L1S might be useful for therapeutic stimulation of both cytokine and cytoxicity-based immune responses . Secretion of IFNγ by NK cells is thought to promote clearance of the bacterial pathogen Francisella tularensis [39] , [40] , [41] . However , this cytosolic intracellular bacterial pathogen normally suppresses innate immune responses [41] , [42] , [43] . We thus hypothesized that boosting of NK cell activation during F . tularensis infection might reduce host susceptibility to this pathogen . To test this hypothesis , we administered purified , LPS-associated L1S or PBS alone by a single i . p . injection 24 hours prior to an i . p . infection with the attenuated live vaccine strain of Francisella tularensis holarctica LVS ( Ft ) . Bacterial burdens in the infected spleens ( Figure 7A ) and livers ( Figure 7B ) were assessed 96 hours post Ft infection . Colony-forming units ( CFU ) recovered from spleens and livers of the L1S treated mice were significantly reduced when compared to the control mice . Consistent with the increase in IFNγ+ NK1 . 1+CD3- cells seen after in vivo L1S stimulation ( Figure 6A , 6B ) , we observed a significant increase in serum IFNγ levels in the mice treated with L1S prior to Ft infection ( Figure 7C ) . To control for the potential effects of LPS associated with purified L1S , we pre-treated mice with LPS or LPS-associated L1S 24 hours prior to Ft LVS infection as above . The CFUs recovered 4 days post-infection were significantly lower in mice pre-treated with LPS-associated L1S compared to LPS alone ( Figure 7D ) . Serum levels of IFNγ were also significantly higher in the L1S versus LPS pre-treated mice ( not shown ) , which correlates with the observed minimal effect of LPS on IFNγ levels in NK cells in vivo ( Figure 6A , 6B ) . These findings suggest that p60 and its derivatives enhance NK cell activation in a biologically relevant manner and may be useful for further development as a therapeutic for immune stimulation .
Bacterial pathogens have developed numerous strategies to interfere with or subvert host immune responses [44] , [45] . Our findings here demonstrated an indirect role for the abundantly secreted L . monocytogenes ( Lm ) p60 protein in modulation of NK cell activity . We showed that Lm secretion of the p60 protein during infection of cultured BMDCs stimulated enhanced activation of naïve NK cells in cell co-culture assays . Moreover , endotoxin-free purified p60 protein was sufficient to stimulate IFNγ production from NK cells in co-cultures containing BMDCs primed with TLR agonists or inflammatory cytokines such as IL-12 . Purified p60 protein bound to the BMDCs and in the presence of priming stimuli this binding correlated with BMDC secretion of the NK cell activating cytokine IL-18 . These findings support the model that p60 indirectly activates NK cells by stimulating a DC surface receptor in a manner that induces secretion of IL-18 . Consistent with this model , IL-18 production by the BMDCs was essential for eliciting IFNγ production by NK cells and cultures of NWNA splenocytes . The known endopeptidase enzymatic activity of p60 was not required for this biological response and stimulation of NK cell activation by p60 or its derivatives was independent of bacterial PGN and muropeptide detection systems dependent on the RIP2 kinase . Rather , the ability to stimulate DC-dependent NK cell responses mapped to an N-terminal fragment of p60 that contains a LysM domain and a bacterial SH3 domain . A polypeptide containing just these domains ( L1S ) was sufficient to stimulate DC-dependent NK cell activation both in cell culture assays and when administered to mice in the absence of Lm infection . These results thus revealed a novel role for a bacterial LysM domain-containing protein in the modulation of mammalian innate immune responses . Our studies here demonstrated that extracellular delivery of p60 protein or the L1S polypeptide in cell culture acted in concert with DCs to stimulate NK cell activation . However , the p60 protein was not sufficient to activate NK cells in the absence of primed BMDCs . In addition , soluble L1S polypeptide triggered NK cell activation when injected into mice without any known mechanism for uptake into the cytosol of host cells . These data suggest that p60 and/or L1S act extracellularly to increase the ability of DCs to promote NK cell activation . Consistent with this interpretation , we observed that p60 protein bound to the surface of BMDCs but not NK cells . Staining of p60 on BMDCs that were fed latex beads suggested that aggregates of p60 are not simply phagocytosed . Furthermore , delivery of p60 into the cytosol of cultured BMDCs using transfection protocols did not improve NK cell activation ( not shown ) . These data suggest that p60 protein acts extracellularly to promote NK cell activation . The fact that infected individuals develop antibodies against p60 further suggests this protein may be abundant extracellularly during Lm infection [31] . Potential sources of extracellular p60 include production by extracellular bacteria , which are known to be present at early and later times of infection [46] , or release of protein upon lysis of infected cells . However , we cannot exclude the possibility that p60 present in the cytosol after phagosomal escape of Lm also contributes to NK cell activation . Indeed , p60 protein is abundant in the cytosol of Lm infected macrophages and stimulates protective cytotoxic T cell ( CTL ) responses [47] , [48] . Since cytokines and TLR agonists are also present during Lm infections , soluble extracellular p60 protein that interacts with DCs or other infected cells during in vivo Lm infection is likely an important stimulus for NK cell activation during in vivo Lm infection . However , our data here ( Figure 1 ) and in a prior publication [17] clearly indicate that there are also p60-independent mechanisms for NK cell activation . The activation of naïve NK cells by DCs infected with live Lm bacteria was previously shown by us and others to require both direct contact between DCs and NK cells and the production of IL-12 and IL-18 [8] , [18] . Lm bacteria obviously contain TLR agonists that can induce IL-12 production to prime NK cell activation during in vivo infection . However , it has not been clear whether specific bacterial factors stimulate IL-18 production and/or cell contact between naïve NK cells and DCs . Our data here implicate the L1S region of p60 as a bacterial factor that promotes IL-18 production by DCs . Specifically , we showed that priming of BMDCs with TLR agonists stimulated IL-12p70 production by these cells and that IL-12p70 could substitute for TLR agonists . In some experiments , we also observed a modest p60-induced enhancement of IL-12 secretion from BMDCs that were already primed with TLR agonists , which is consistent with the ability of IL-18 to positively regulate IL-12 production . However , neither TLR agonists nor IL-12 were sufficient to stimulate NK activation in the absence of p60 protein and the IL-18 production elicited by p60 . Moreover , despite the presence of IL-12 and IL-18 , stimulation of BMDCs with TLR agonists and p60 was insufficient to stimulate NK cell activation when there was not direct cell-cell contact between the BMDCs and the NK cells . The p60 treatment appeared to induce a more activated phenotype in BMDCs but it did not alter the expression by BMDCs of several known ligands for NK cell activating and inhibitory receptors . Thus , there exist at least three possible explanations for the contact requirement: ( 1 ) The p60 stimulation triggers both IL-18 secretion and expression of an activating ligand by the BMDCs . This ligand is not one we have tested and may be novel . ( 2 ) Contact merely serves to increase the local concentration of IL-18 ( and perhaps IL-12 ) above some threshold that normally prevents activation of the naïve NK cells . This may be facilitated by immunological synapses formed between the DC and NK cells , as previously suggested [13] , [14] . ( 3 ) BMDCs constitutively express ( or are induced to express e . g . by p60 or IL-12 ) a surface associated “co-stimulatory” factor that is required to “prime” the NK cells for responsiveness to IL-18 . Ongoing and future studies focused on identification of putative ligands or co-stimulatory factors may resolve which , if any , of these possible explanations is correct . NK cells are the major source of IFNγ production early after viral and bacterial infections . IFNγ normally plays a protective role in immunity to Lm and other pathogens , including F . tularensis ( Ft ) . IFNγ induces CD4 Th1 differentiation , stimulates cytotoxic CD8 cells , and activates macrophages to become more bactericidal [49] , [50] . During Ft infection , IFNγ-positive NK cells are quickly recruited to sites of infection , where they promote granuloma formation and limit bacterial spread [39] , [41] . We found that injection of L1S polypeptide into mice was sufficient to activate NK cells to produce IFNγ , particularly at the site of injection . We also found increased serum levels of IFNγ persisting through infection in mice pre-treated with L1S polypeptide . Presumably , the IFNγ produced by these NK cells created a non-permissive environment for Ft expansion . Thus , when Ft was inoculated at the same site as the L1S polypeptide , its growth was significantly reduced compared to inoculations in the absence of L1S . It will be important to determine whether L1S polypeptide injection might also protect against other routes of Ft infection and against other pathogens . In contrast to Ft infection , the results of in vivo depletion studies suggest that NK cells are associated with increased susceptibility of mice to Lm [17] , [51] , [52] , [53] , [54] and the expression of p60 by Lm increased host susceptibility to systemic Lm infection [17] , [21] . Thus , production by Lm of a protein that promotes NK cell activation correlates with the fact that NK cell activation increases susceptibility to Lm . It was also previously reported that IFNγ production by NK cells fails to protect mice against systemic Lm infections [55] . This may be due to suppression of macrophage responsiveness to IFNγ during early stages of Lm infection [56] . Thus , Lm produces a protein that enhances NK cell activation and also has been shown to be more pathogenic in the presence of NK cells . It will thus be of interest in future studies to understand the mechanisms by which activated NK cells promote Lm pathogenicity . In contrast to Lm , Ft normally suppresses host inflammatory responses during the initial stages of infection [42] , [43] . The Ft genome contains several LysM-containing proteins , but using BLAST searches we failed to identify any Ft proteins whose LysM-domains showed more than 20% identify to the LysM1 region of p60 . Thus , it is possible that the LysM proteins present in Ft have evolved to lack residues critical for binding to DCs or activation of IL-18 secretion by DCs . Consistent with this model , we found that no IFNγ was produced by NWNA splenocytes cultured with Ft-infected BMDCs ( data not shown ) . However , this issue will need to be further investigated , since it is also possible that Ft LysM proteins are not secreted and thus accessible to bind DCs in the same manner as the Lm p60 L1S region . NK cells are attractive targets for therapy in cancer and infectious diseases as they can directly kill target cells . NK cells also regulate immune and autoimmune B cell and T cell responses through production of IFNγ or inhibitory cytokines such as TGFβ and IL-10 [57] , [58] . NK cells have additionally been shown to impact Type I diabetes , multiple sclerosis , and other diseases associated with inflammation [59] , [60] . Our findings demonstrated use of the p60 protein to stimulate activation of cultured NK cells . L1S also demonstrated effective NK cell activation when administered in vivo . With refinement , p60 or L1S may be adapted to therapeutic use to harness anti-cancer or immune regulatory effector mechanisms of NK cells . Further experimentation on the clinical and biological effects of p60 protein may thus provide novel approaches to manipulate host immune responses . Additionally , it will be of interest to determine whether and how LysM-containing proteins from other pathogens modulate innate immune responses . Such studies should improve our understanding of bacterial pathogenesis and the role of NK cells in immune responses .
This study was carried out in strict accordance with the recommendations of the Public Health Service Policy on the Humane Care and Use of Laboratory Animals , the Guide for the Care and Use of Laboratory Animals , and the Association for Assessment and Accreditation of Laboratory Animal Care . The protocols used were approved by the Institutional Animal Care and Use Committee at National Jewish Health ( Protocol Permit AS2682-9-13 ) . All efforts were made to minimize suffering . C57BL/6 and B6 . IL-18-/- mice were obtained from Jackson labs . Breeders of B6 . Rip2-/- mice were generously provided by K . Kobayashi ( Dana-Farber/Harvard , Boston , MA ) . Mice were bred and housed in the Biological Research Center of National Jewish Health . Studies were performed with the approval of the National Jewish Health Institutional Animal Care and Use Committee . Wild type Listeria monocytogenes 10403s was used in these studies . In-frame deletion of p60 in 10403s was done by allelic exchange , as described [21] . The full p60 complementation mutant expresses a secreted His-tagged p60 protein expressed from the pPL2-derived vector pIMK2 , a generous gift from C . G . M . Gahan described in [61] . The ΔLysM1-p60 complementation mutant lacks the first LysM1 domain , residues 26-69 , and is also expressed from the pIMK2 vector . SOE PCR primer sequences are provided in Table S1 . 10403s Δp60 was transformed with the His-p60 construct or ΔLysM1and p60 protein secretion was assayed by immunoblot of TCA precipitated of supernatants from overnight Lm cultures . Plasmid DNA encoding the ΔLD-p60 , lacking residues 138-179 , was provided by E . Pamer ( Sloan-Kettering , NY ) and described in [62] . The mutated gene was amplified with primers described in Table S1 and subcloned into the pPL-2 vector for transformation into 10403s Δp60 Lm . The Francisella tularensis live vaccine strain ( LVS ) holarctica type b was obtained from ATCC BEI Resources ( Manassas , VA ) . Escherichia coli TOP10 cells were obtained from Invitrogen ( Carlsbad , CA ) and were used to clone and express all His-tagged purified proteins in this study . Femoral bone marrow was flushed and cultured in RPMI 1640 ( high glucose ) ( Gibco , Invitrogen ) with 10% FBS , . 1% betamercapto-ethanol , 1%L-glutamine , 1% sodium pyruvate , 1% penicillin/streptomycin , and 2 ng/ml GM-CSF . BMDC were washed on days 2 and 4 , and harvested on day 7 . 3×105 cells were plated per well of a 24-well plate in triplicate for >12 hours in antibiotic-free media , then infected with log phase 10403s wt or Δp60 at MOI of 1 for 1 hour . Cells were then washed and treated with 10 µg/ml gentamycin . For protein stimulation , 3×105 BMDC were treated with 10 µg purified protein plus or minus pre-treatment with 10 ng/ml ultra-pure LPS , 10 ng/ml mono-phosporo-Lipid A ( MPA ) ( Sigma-Alderich , St . Louis , MO ) , or 20 µg/ml Polyinosine-polycytidylic acid ( PIC ) ( Invivogen , San Diego , CA ) for 3 hours . Splenocytes were prepared and enriched for lymphoctyes by nylon wool non-adherence ( NWNA ) as described [8] . Lymphocytes were 5-6% CD3- NK cells based on staining with NK1 . 1 ( PK136 ) and CD3 ( 145-2C11 ) ( BD Biosciences Franklin Lakes , NJ and eBioscience San Diego , CA ) . The splenocytes were added to the BMDC at a 0 . 1∶1 NK cell∶BMDC ratio at 2 hours post-infection . To obtain purified NK and T cells from NWNA splenocytes , cells were stained with NK1 . 1 and CD3 and sorted by flow cytometry on the Synergy ( Icyt , Champaign , IL ) . Purified NK1 . 1+/CD3- NK cells ( 3×104 ) , CD3+/NK1 . 1- T cells ( 5×104 ) and NK1 . 1-/CD3- cells ( 3×104 ) were added to 3×105 BMDC per well . To test NWNA splenocytes activation in the absence of BMDC , a 50% bead slurry of Ni-NTA agarose beads ( Invitrogen ) was washed 5 times with PBS , associated with 50 µg L1S/well , washed 2 times with PBS , and then was added to NWNA splenocytes in the presence or absence of BMDC . DNA coding for the mature p60 , p60C389A , Lm 0394 , Np60 , Cp60 , and L1S were cloned into the pTrcHis-TOPO TA cloning vector ( Invitrogen , Carlsbad , CA ) for IPTG-induced expression in TOP 10 E . coli . Primers are listed in Table S1 . The phage autolysin HPL511 was purified from a construct supplied by M . Loessner ( Zurich ) . E . coli were lysed with BugBuster ( Novagen , Gibbstown , NJ ) in 20 mM Na phosphate , 0 . 5 M NaCl , and 20 mM imidazole , pH 7 . 4 , containing protease inhibitor and 2 mg/ml lysozyme . Proteins were purified using HisTrap FF 5 ml affinity columns ( GE , Piscataway , NJ ) on an Akta FPLC ( GE ) . Further purification was achieved with Hi-Trap FF or HP ( GE ) cationic exchange in 50 mM HEPES buffer . LPS was removed from the proteins using polymyxin B columns as indicated by the manufacturer ( Thermo Scientific , Waltham , MA ) . Supernatant levels of murine IFNγ , IL-12p70 , and IL-18 were measured at 21 hours post-infection using commercial ELISA kits ( BD Biosciences , MBL International , Woburn , MA ) . BMDCs ( 3×105 per coverslip ) were treated with 30 µg/ml purified p60 with or without 1×108 FITC-labeled 0 . 5 um latex beads ( Polysciences , Inc , Warrington , PA ) . p60 was probed with PFII rabbit anti-p60 ( supplied by E . Pamer , New York ) and Fab ( ab′ ) 2 goat-anti-rabbit Cy3 ( Invitrogen ) . Actin was visualized with Alexa-488 or Alexa-680 phalloidin and nuclei were stained with DAPI ( Invtrogen ) . Slides were viewed with the Leica DMRXA ( Leica Microsystems Inc . , Bannockburn , IL ) . Data were collected at 100x and 40x magnification in oil at room temperature . Lenses were 100x oil , numerical aperture 1 . 4- 0 . 7 , and 40x oil numerical aperture 1 . 25-0 . 75 . Images were taken using the Coolsnap XQ camera ( Photometrics , Tucson , AZ ) and processed with Slidebook 5 ( Intelligent Imaging Innovations , Inc . , Denver , CO ) . Minimal contrast adjustment was applied equally to experimental and control merged images . Images were sized and annotated using Photoshop ( Adobe Systems , Inc . , San Jose , CA ) . BMDCs were plated in triplicate and primed for 3 hours with 30 ng/ml LPS and then treated with 30 µg/ml purified L1S p60 protein-derived peptide for 4 hours . The cells were then lifted and surface stained for Kb ( AF6-88 . 5 . 5 . 3 ) , Db ( 28-14-8 ) , MHC-II ( M5/114 . 15 . 2 ) , RAE1γ ( CX1 ) , RAE1δ ( RD-41 ) , MULT-1 ( 5D10 ) , CD229/Ly9/SLAMF3 ( Ly9ab3 ) , Ly-108/SLAMF6 ( eBio13G3-19D ) , CD150/SLAMF1 ( 9D1 ) , CD84/SLAMF5 ( mCD84 . 7 ) , and CD48/SLAMF2 ( HM48-1 ) . All antibodies were from eBioscience ( San Diego , CA ) . Cells were run on a LSRII ( BD Biosciences ) and 50 , 000 events were collected . FlowJo software ( Tree Star Inc , Ashland , OR ) was used to analyze samples . 10 µg each of p60 , p60C389A , and 0 . 25 µg of phage autolysin HPL511 were loaded into native 7 . 5% PAGE gels with . 02% heat-killed Lm as PGN substrate . The gels were re-natured in 25 mM Tris ph 7 with 1 mM DTT and 10 mM CaCl2 , shaking overnight at 37°C . Zymography activity was visualized by staining with 0 . 01% methylene blue in 0 . 1%KOH . Female mice between ages 8–10 weeks were treated intraperitonally with 500 µg purified L1S or 10 ng/ml LPS in 300 µl 0 . 2 M sodium phosphate buffer . For NK cell IFNγ intracellular staining , peritoneal infiltrates were harvested by injecting the peritoneum with 10 ml ice cold PBS with 5 mM EDTA . After light shaking , the fluid was recovered , and cells were stained as described below . Spleens were harvested at 24 into RPMI 1640 ( Gibco , Invitrogen ) . Spleens were treated with 1 mg/ml collagenase in Hank's Buffered Salt Solution ( HBSS ) plus cations ( Invitrogen , Carlsbad , CA ) for 30 minutes , mashed through a cell strainer into a single cell suspension and treated with RBC Lysis Buffer ( 0 . 15 M NH4Cl , 10 mM KHCO3 , 0 . 1 mM Na2EDTA , pH 7 . 4 ) and stained as described below . Splenocytes and Peritoneal infiltrates were counted and 2×106 cells were incubated in RP-10 media ( RPMI 1640 , 10% FBS , 1% L-glutamine , 1% Sodium Pyruvate , 1% Penicillin , 1% Streptomycin and 0 . 1% β-mercaptoethanol ) plus GolgiPlug ( BD Biosciences , Franklin Lakes , NJ ) for 3 hours . Cells were then incubated in anti-CD16/32 ( 2 . 4G2 hybridoma supernatant ) to block Fc receptors . Surface staining was performed first and included anti-CD3 ( clone 145 2C11 ) and anti-NK1 . 1 ( clone PK136 ) . Cells were then fixed and permeabilzed in a 4% paraformaldehyde and saponin solution and stained with anti-IFNγ ( clone XMG1 . 2 ) and anti-granzyme B ( 16G6 ) ( eBioscience , San Diego , CA ) . Cells were run on a LSRII ( BD Biosciences ) and 100 , 000 events were collected . FlowJo software ( Tree Star Inc , Ashland , OR ) was used to analyze samples . Splenocytes from co-culture experiments were collected 10 hours post-infection , cultured with GolgiPlug ( BD Biosciences ) for 3 hours , and stained as above . BMDCs ( 3×104 per well ) were treated or not with 10 ng LPS with or without 10 µg L1S per well for 2 hours . NK-enriched NWNA splenocytes were added to the BMDCs at 2 hours as described in NK-activation and Co-culture . After 21 hours of co-culture , the NWNA splenocytes were collected from co-culture , counted , and added to 5×104 B16F10 mouse melanoma cells ( ATCC , Manassas , VA ) at Effector∶Target ratios of 1∶10 , 1∶10 , and 10∶1 based on the estimated number of NK cells in the NWNA splenocytes ( 5% ) . The effector and target cells were incubated for 4 hours and cytotoxicity based on LDH release was measured using the Cytotox96 cytotoxicity kit as per manufacturer instructions ( Promega , Madison , WI ) . 6–8 week old female mice were pre-treated with 300 µl PBS alone or with 500 ng LPS with or without 500 µg purified L1S , injected i . p . After 24 hours , the mice were infected i . p . with ∼104 LVS strain of F . tularensis ssp . holarctica LVS ( Ft ) . Livers and spleens were harvested at 96 hours post Ft infection into 0 . 02% Nonidet P-40 . Livers and spleens were homogenized in a protected fume hood for 1 minute and 2 serial dilutions of homogenate were plated on BHI ( Brain and Heart Infusion ) ( BD Biosciences ) agar plates . Plates were incubated at 37°C , 7 . 5% CO2 with humidity for 72 hours and colonies were counted to determine colony forming units per organ . Serum levels of IFNγ were measured by ELISA . Statistical analysis was performed using Graph Pad Prism 5 ( La Jolla , CA ) . P values were assessed using unpaired , two-tailed Student's t tests ( α = 0 . 05 ) . In the figures , * denotes P values between 0 . 05 and 0 . 01 , ** denotes P values between 0 . 01 and 0 . 001 , and *** denotes P values < or = 0 . 001 . p60 ( NCBI accession ZP_05235088 . 1 ) , Lm 0394 ( NCBI accession ZP_05235264 . 1 ) .
|
Pathogens have evolved diverse strategies to influence host immune responses . By studying these strategies , we may learn how to therapeutically intervene to manipulate immune responses during infectious and other diseases . In this study , we investigated how the bacterial pathogen Listeria monocytogenes ( Lm ) stimulates activation of an innate immune cell type called the natural killer ( NK ) cell . NK cells protect against certain infections , tumors , and autoimmune diseases , but appear to play a deleterious role in the context of Lm infection . We found that putative carbohydrate and protein interaction domains of a heavily secreted Lm protein and virulence factor , p60 , indirectly stimulate NK cells both during infection and in the absence of other bacterial factors . Treatment of mice with this region of p60 stimulated NK cell activity and was protective in a mouse model of systemic infection by an NK cell sensitive bacterial pathogen , Francisella tularensis . These studies suggest that derivatives of p60 protein may prove to be useful tools for activation of NK cells and demonstrate therapeutic use of this bacterial immune modulating factor .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"gram",
"negative",
"immune",
"cells",
"nk",
"cells",
"immunity",
"immune",
"activation",
"innate",
"immunity",
"immunology",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"bacterial",
"pathogens",
"gram",
"positive"
] |
2011
|
A LysM and SH3-Domain Containing Region of the Listeria monocytogenes p60 Protein Stimulates Accessory Cells to Promote Activation of Host NK Cells
|
Staphylococcus aureus is a major human pathogen that causes a range of infections from acute invasive to chronic and difficult-to-treat . Infection strategies associated with persisting S . aureus infections are bacterial host cell invasion and the bacterial ability to dynamically change phenotypes from the aggressive wild-type to small colony variants ( SCVs ) , which are adapted for intracellular long-term persistence . The underlying mechanisms of the bacterial switching and adaptation mechanisms appear to be very dynamic , but are largely unknown . Here , we analyzed the role and the crosstalk of the global S . aureus regulators agr , sarA and SigB by generating single , double and triple mutants , and testing them with proteome analysis and in different in vitro and in vivo infection models . We were able to demonstrate that SigB is the crucial factor for adaptation in chronic infections . During acute infection , the bacteria require the simultaneous action of the agr and sarA loci to defend against invading immune cells by causing inflammation and cytotoxicity and to escape from phagosomes in their host cells that enable them to settle an infection at high bacterial density . To persist intracellularly the bacteria subsequently need to silence agr and sarA . Indeed agr and sarA deletion mutants expressed a much lower number of virulence factors and could persist at high numbers intracellularly . SigB plays a crucial function to promote bacterial intracellular persistence . In fact , ΔsigB-mutants did not generate SCVs and were completely cleared by the host cells within a few days . In this study we identified SigB as an essential factor that enables the bacteria to switch from the highly aggressive phenotype that settles an acute infection to a silent SCV-phenotype that allows for long-term intracellular persistence . Consequently , the SigB-operon represents a possible target to develop preventive and therapeutic strategies against chronic and therapy-refractory infections .
S . aureus is a major human pathogen that can infect almost every organ in the body and cause destructive infections [1] . Besides tissue damage the ability to develop persisting and therapy-refractory infections poses a major problem in clinical practice , such as endovascular and bone infections [2 , 3] . Chronic infections require prolonged antimicrobial treatments and can have a dramatic impact on the patients`quality of life , as they often afford repeated surgical interventions with the risk of amputation or loss of function [2 , 4] . To induce an infection S . aureus expresses a multitude of virulence factors , including surface proteins and secreted components , like toxins and peptides [1] . Toxins and other secreted factors are mainly directed against invading immune cells , but can also cause tissue damage that enables the bacteria to enter deep tissue structures [5 , 6] . Yet , S . aureus is not only an extracellular pathogen , but can also invade a wide variety of mammalian cells , such as osteoblasts , epithelial- and endothelial cells [7–9] . All cells discussed in the literature as “non-professional phagocytes” possess mechanisms that nevertheless permit endocytotic uptake and degradation of microorganisms [10–12] . To evade the intracellular degradation machineries the bacteria have evolved different strategies , such as the killing of host cells or the escape from the lysosomal compartments and silent persistence within the intracellular location [8 , 13 , 14] . Only recently , we demonstrated that S . aureus can dynamically switch phenotypes from a highly aggressive and cytotoxic wild-type form to a metabolically inactive phenotype ( small colony variants , SCVs ) that is able to persist for long time periods within host cells without provoking a response from the host immune system [9] . In their intracellular location the bacteria are most likely very well protected from antimicrobial treatments and the host´s defense system . This is a possible reservoir for chronic and recurrent infections . Yet , the environmental changes encountered by invading bacteria during the passage from an extracellular to the intracellular milieu and during long term persistence within the intracellular shelter most likely cause diverse stress conditions . The adaptation mechanisms involved and how bacteria cope with this stress , are largely unknown , but probably involve global changes in gene expression to promote survival . It is well known that S . aureus possess a large set of regulatory factors that control the expression of virulence determinants [15–18] . A very important and well-studied system is the accessory gene regulator ( agr ) -locus with the effector molecule RNA III [19 , 20] . Many S . aureus derived factors that induce inflammation or cell death are under the control of this system . Important cytotoxic components regulated by the agr-system are the pore-forming α-hemolysin ( α-toxin , Hla ) and the phenol-soluble modulins ( PSMs ) , which are strong cytotoxic and pro-inflammatory factors in different host cell types [6 , 21–23] . SarA , which is the major protein encoded by the sar-locus , is believed to contribute to the activation of agr expression [24 , 25] . This is supported by findings of many S . aureus infection models demonstrating that mutations of either loci result in attenuation of virulence [26–28] . Beyond that , it has been shown that SarA influences the regulation of several virulence factors independently of agr , e . g . expression of adhesins [25 , 29 , 30] . The alternative sigma factor B ( SigB; σB ) modulates the stress response of several Gram-positive bacteria , including S . aureus [31–33] . SigB is responsible for the transcription of genes that can confer resistance to heat , oxidative and antibiotic stresses [16 , 31 , 34 , 35] . The sigB system is linked to the complex S . aureus regulatory network , as it increases sarA expression , but decreases RNA III production [36] . During the infection process the bacteria encounter different stressful conditions that they need to overcome in order to settle and maintain an infection . In the acute infection they have to fight against invading immune cells and destroy tissue cells to enter deep tissue structures , whereas during the chronic phase a major challenge is most likely the hostile intracellular location deprived of nutrients and lysosomal degradation . Consequently , the host-pathogen interaction needs to be very dynamic . Yet , to date no studies have followed the adaptation mechanisms and the impact of regulators during the whole infection process . In this work we focused on the interaction of the global regulatory systems agr , SarA and SigB and we demonstrate a crucial function for SigB in bacterial adaptation mechanisms and the formation of SCV-phenotypes .
For our study we generated single mutants for the functions of SigB ( ΔsigB ) , agr ( Δagr ) , SarA ( ΔsarA ) and the complemented mutant for sigB ( ΔsigB compl . ) , three double mutants for SigB , agr , SarA ( Δagr/ΔsarA , ΔsigB/Δagr , ΔsigB/ΔsarA ) and a triple mutant ( ΔsigB/Δagr/ΔsarA ) in S . aureus LS1 , a strain derived from mouse osteomyelitis [37] . Several mutants were also generated in the rsbU complemented derivative of the laboratory strain 8325–4 , SH1000 [33] ( Supp . S1 Table ) . All strains and mutants were characterized by growth curves , which did not reveal any substantial differences between mutants and parent strains ( S1 Fig ) . Yet , differences in hemolysis were present in many mutants , particularly in the agr- , sarA- , double and triple mutants , which exhibited low hemolysis ( S1C Fig ) . Furthermore , the strain LS1 and the corresponding mutants were analyzed by LC-MS/MS mass spectrometry , which was focused onto the culture supernatants to provide an overview on the levels of virulence factors in each strain ( Fig 1A and S3 Table ) . These data show that particularly the sarA-mutant and even more the double and the triple-mutants released a much reduced number of virulence factors associated with disease development compared with the wild-type strain LS1 ( reduced levels of virulence factors associated with disease; deep green areas , Fig 1A ) ; in particular different adhesive proteins and toxins were present in reduced levels in culture supernatants ( Fig 1B ) . Yet , for the FnBPs that are important for host cell invasion we detected similar levels compared with the wild-type LS1 in most mutants that can account for the invasive capacity of the strains in host cells ( S5 Fig ) . Most importantly , the sigB-mutant showed only modest alterations in protein levels ( Fig 1A ) , but increased levels of α-toxin that is known as a strong proinflammatory and cytotoxic factor after host cell invasion ( hla , Fig 1B ) . The other listed toxins , e . g . lukD and lukE , are reported as non-hemolytic and only poorly leucotoxic toxins [37] . As secreted virulence factors are particularly directed against professional phagocytes , we tested the effect of bacterial supernatants on neutrophils ( PMNs ) isolated from humans and mice . In line with the proteomic data , only the strains with high levels of virulence factors ( LS1 , ΔsigB , ΔsigB compl . ) caused cell death , whereas all other mutants with reduced levels of toxins induced significantly less cytotoxicity ( Figs 2A and 2B and S2A–S2C ) . This effect was concentration dependent and revealed highest levels of cell death in response to supernatants of the sigB-mutant ( S2C and S2D Fig ) . Next we analyzed levels of chemokine expression in cultured tissue cells , such as osteoblasts and endothelial cells , 48 h after infection by real time PCR ( Fig 2C and 2D and S4C and S4D Fig ) and 24 h after infection by ELISA-measurements ( S3C and S3D Fig ) . In contrast to the wild-type strain , all double-and triple-mutants ( including mutations in SigB ) exhibited reduced cell activating activity , whereas the sigB-single-mutant often caused even more cell activation than the wild-type strain . Furthermore , these effects were independent of the bacterial background and of the infected host cell types , as they were reproduced with selected mutants generated in strain SH1000 ( S3D Fig ) . All effects were equally present in bone and endothelial cells ( S4C and S4D Fig ) regardless of the observation that endothelial cells took up higher amounts of bacteria than osteoblasts ( S4A Fig ) . Nevertheless , endothelial cells expressed in general lower levels of chemokines than osteoblasts ( S4B Fig ) . Taken together , these results suggest that the agr- and SarA-systems are required to mount an aggressive and cytotoxic phenotype during acute infection , while SigB appears to have a restraining function on virulence . Nevertheless , since double and triple mutants are weak in virulence , the interaction of SigB with the agr- and SarA-systems is not clear during acute infection . To test the function of the global regulatory systems in the course from acute to chronic infection , we infected osteoblast and endothelial cell cultures with wild-type and mutant strains and analyzed their ability to persist intracellularly for 9 days . All strains were invasive in osteoblasts to a similar extent ( S5 Fig ) and induced cell death ranging around 50% immediately after infection ( S6A and S6B Fig ) . Yet , in the following 2–3 days the integrity of the infected cell monolayers were fully recovered and the rate of cell death was reduced to control levels ( S6C Fig ) . In general the numbers of intracellular bacteria were decreased during the whole infection course ( Figs 3A and S7A ) , but considerable differences between the strains appeared after several days ( 9 days , Figs 3B and S7B ) . The agr/sarA- and the sigB/sarA-double mutants as well as the triple mutant were able to persist within the intracellular location at significantly higher numbers ( up to 100-fold ) than the corresponding wild-type strain ( Fig 3A and 3B ) . By contrast , the sigB-mutants were completely cleared from the host cells within 7–9 days , whereas this effect could be fully reversed by the complementation of sigB ( Figs 3A and S7A ) . To test whether this effect is specific for sigB-mutants , we further tested mutants for other virulence or regulatory factors such as for sae and hla which did not reveal any differences in the numbers of intracellular persisting bacteria compared with the wild-type strain ( S8 Fig ) . From our previous work we know that bacterial persistence is associated with dynamic SCV-formation . As recently described [9] we discovered an increased rate of SCV-formation after several days of intracellular bacterial persistence , whereby the recovered SCV were not stable , but the majority reverted back to the wild-type phenotype upon 2 to 5 subcultivating steps on agar plates . Interestingly , in the present study we found that all sigB-mutants completely failed to develop SCV phenotypes after 7 days of intracellular persistence ( Fig 3C and 3D and S7C and S7D Fig ) . By analyzing the recovered colonies from sigB-mutants , we observed much less phenotypic diversity than in the wild-types and other mutants , as the plates revealed only uniform large white colonies . Again these effects could be reversed by complementation of the sigB-mutations with an intact sigB-operon , thus proving a clear and specific connection between the bacterial ability to form dynamic SCVs and the SigB-system . To further explain the differential ability of the mutants to persist , we evaluated the expression of the global regulatory systems during the long course of the infection . To accomplish this , we extracted RNA from infected host cells ( HUVEC; as they can host higher numbers of bacteria , S4A Fig ) at day 2 and day 7 p . i . and measured the expression of agr , sarA and sigB and the related factors hla ( regulated by agrA ) , aur ( repressed by sarA [38] ) and asp23 ( regulated by sigB [39] ) in the wild-type strain and corresponding mutants by quantitative real-time PCR ( Fig 4 ) . As expected high levels of both agrA and sarA were only expressed by the strains LS1 , ΔsigB and ΔsigB compl . that resulted in high levels of hla expression in the acute phase of the infection . By contrast , the agr/sarA-double mutant expressed sigB and asp23 at significant higher levels than the wild-type strain during chronic infection ( Fig 4 ) and was able to form higher numbers of SCV phenotypes ( Figs 3C and S7C and S7D ) . Taken together , our results show that a concomitant downregulation of agrA and sarA promotes long-term intracellular persistence of S . aureus . SigB promotes chronic infections and is highly associated with the bacterial ability to form SCVs . Only recently , phagosomal escape to the cytoplasm was reported for different S . aureus strains early after host cell invasion [13 , 40] . In a further approach we analyzed whether an early phagosomal escape is a prerequisite for persistence . Therefore we used a reporter recruitment technique based on the host cell line A549 genetically engineered to produce a phagosomal escape marker [13] . Within the first 2 h after host cell infection we detected phagosomal escape for the wild-type strain LS1 , the sigB- , the sigB compl . - and the sigB/agr-mutants ( Fig 5A–5D ) . These strains showed only weak changes and down-regulation of virulence factor expression by proteomic analysis compared with the wild-type ( Fig 1A ) and were not able to persist at high bacterial numbers ( Fig 3B ) . As the single agr and sarA mutants readily lost their ability to translocate to the cytoplasm , apparently both agr- and sarA-regulated factors are required for the escape mechanism . The activity of SarA alone is sufficient only in case of a non-functional SigB-system , indicating a modulating role of SigB in virulence factor expression . Our results show that an early phagosomal escape is not required for persistence . Further on , mutants that persisted at high bacterial numbers did not escape to the cytoplasm . As phagosomal escape could not be detected at later stages of infection ( up to 24 h; as shown in for the triple mutant; Fig 5E ) , it must be assumed that these mutants are not degraded within phagolysosomes and thus persist in phagosomes at high bacterial numbers . To test the ability of the wild-type and the mutant strains to establish an infection in vivo , we next performed experiments using a rat localized osteomyelitis model [41] , where defined numbers of bacteria ( 1x106 CFU; strain SH1000 and mutants ) were directly injected into the bones ( Fig 6 ) . Using this model we aimed to study the complex interaction of the bacterial strains with the immune response of the host organism . After 4 days ( acute stage ) and 14 weeks ( chronic stage ) groups of rats were sacrificed and the tibial bones were used for histology or morphometric analysis ( osteomyelitis index; OI , Fig 6F ) to assess for infection severity . After 4 days and 14 weeks the bacterial loads were determined by quantitative culture of bone homogenates . The data clearly demonstrated that the single mutants ΔsigB , Δagr and ΔsarA were found in lower numbers within the bones and caused less inflammation than the wild-type strain ( Fig 6B–6E ) . We further tested as a selective double mutant the agr/sarA double mutant that persisted in high numbers within cells in culture experiments ( Figs 3 and S7 ) . By contrast , in the animal model we detected drastically reduced CFU and a lower osteomyelitis index in rats challenged with the agr/sarA double mutant when compared with rats infected with the parental wild-type strain . The agr/sarA-double mutant was found almost as avirulent as the low-pathogenic Staphylococcus carnosus strain TM300 [42] , which lacks most S . aureus virulence factors ( Fig 6B and 6C ) . Nevertheless , histological analysis 4 days p . i . revealed that in all experimental groups and control rats the bones were densely infiltrated with immune cells after S . aureus challenge ( Fig 6A ) , indicating that both the wild-type and the mutant strains attract immune cells and sustain the infection . Only the wild-type strain , however , was able to settle and replicate at high bacterial numbers , to induce severe bone destruction and to develop into chronicity . These findings supports the hypothesis that regulators agr , SarA and SigB need to be functional to enable S . aureus to successfully survive during the whole infection process .
S . aureus is one of the most frequent causes of osteomyelitis and endovascular infections that often take a therapy-refractory or chronic course . Many clinical studies show that persistent infections are highly associated with the SCV phenotype [4 , 43] . Only recently , we were able to demonstrate that the formation of dynamic SCVs is an integral part of the long-term infection process that enables the bacteria to hide inside host cells , but also to rapidly revert to the fully-aggressive wild-type phenotype , when leaving the intracellular location and causing a new episode of infection [9] . Therefore , the pathogen-host interaction must be very dynamic and most likely requires global transcriptional changes on the bacterial side to promote survival of the pathogen . This study was aimed at detecting regulatory factors that mediate this dynamic infection and adaptation strategies . To this purpose , we focused on a set of important S . aureus regulators/regulatory loci agr , SarA , and SigB that are linked together in a global regulatory network . Each one of these regulators/regulatory loci is involved in the control of the expression of many virulence factors such as adhesive and cytotoxic components . For our experiments we used various in vitro and in vivo infection models to analyze the impact and dynamics of the regulators from acute to chronic infection . In models of acute infection we demonstrated that active agr- and SarA-systems are required to cause inflammation and cytotoxicity . Our results are in line with many published infection models , showing that both factors contribute to disease development [27 , 44–46] . We further confirmed that strains become almost avirulent , when both factors are inactive . In case of a single mutation , agr and sarA may partly compensate for each other [47] , but the compensation is not sufficient in all functions , e . g . , in phagosomal escape ( Fig 5 ) and in in vivo infections ( Fig 6 ) . In the acute stage of infection the bacteria need to express a set of virulence factors , including toxins and exoenzymes , to fight against recruited immune cells [48] and to destroy and invade deep tissue structures at high bacterial numbers . Particularly toxins and cytotoxic factors are under the tight control of agr and sarA [19 , 47 , 49] . The role of SigB in acute infection is less clear . Inactivation of sigB in S . aureus has been reported to decrease infectivity in some murine infection models [50–52] , but was ineffective in others [53 , 54] . In line with these conflicting findings , we observed on the one hand that single mutants in sigB express higher levels of toxins , e . g . , α-toxin , and become even more virulent ( Fig 2C and 2D and S2C and S2D Fig ) suggesting a moderating role of sigB on the expression of secreted proinflammatory factors . On the other hand double mutants lacking sigB were almost avirulent ( Fig 2C ) indicating a further function of SigB within the regulatory network during acute inflammation . This is supported by recent work showing a fast and transient upregulation of sigB in the first hours following host cell invasion and the requirement of SigB for early intracellular growth [55] . In the longer course of the infection bacteria can be situated within host cells , like in the cell culture models . As almost all types of host cells contain killing and clearing machineries [10] , the persisting bacteria need to develop mechanisms to resist degradation that can be achieved by two different pathways according to the results from our cell culture experiments: Certain mutants that reveal significant metabolic changes including downregulation of virulence factors ( Fig 1 ) , and do not escape from the phagosomes after host cell invasion ( Fig 5 ) can persist within their host cells partly at higher numbers than the wild-type phenotype ( Fig 3 ) . An example of this is the triple mutant ∆sigB∆agr∆sarA that fails to form SCVs , but is largely avirulent and can persist at high bacterial numbers . Recent work demonstrated that phagosomal escape is largely dependent on the agr-regulated phenol-soluble modulins ( PSMs ) [6 , 13] , but further agr or sarA regulated factors could also be involved , as single mutants in agr or sarA were already compromised in their escape [56–58] . This mechanism of persistence is restricted to strains that lack expression of agr and/or sarA-regulated virulence factors . Our results indicate that these strains are less prone to degradation and can “passively” persist inside host cells , possibly within their initial phagosomes after host cell invasion even without forming SCVs . Further on , persistence is also possible when bacteria express virulence factors and escape from their phagosomes to the cytoplasm . Persistence obviously requires an adaptation to the intracellular environment that could be attributed to the function of SigB during the long course of the infection . SigB is an important staphylococcal transcription factor that is associated with various types of stress-responses [31 , 33 , 35 , 59] , was shown to be upregulated in stable clinical SCVs and was associated with increased intracellular persistence [60] . According to our results SigB is also involved in stress-resistance that promotes “active” intracellular survival during long-term persistence: The first important finding is that ΔsigB-mutants were not able to persist , as they were completely cleared by their host cells within 9 days ( Fig 3B ) . Additionally , the agr/sarA-double mutants that were found at the highest numbers during long-term persistence ( Figs 3B and S7B ) displayed the highest levels of sigB expression ( Fig 4 ) . Finally , a result of major significance is that SigB is required for dynamic SCV formation , as all mutants deficient in sigB were not able to form SCV-phenotypes ( Figs 3C and S7C ) and agr/sarA-double mutants that highly expressed sigB ( Fig 4 ) developed the highest levels of SCVs . Only recently SigB was described as an important virulence factor in stable SCVs that mediates biofilm formation and promotes intracellular bacterial growth [61] . Yet , in our work the recovered SCVs were not stable , but rapidly reverted back to the wild-type phenotype upon subcultivation . Consequently , we describe the formation of dynamic SCVs for persistence as an additional central function that is dependent on an intact SigB-system . Taken together , our results demonstrated that intracellular bacterial persistence is promoted by the silencing of agr- and sarA-regulated factors and/or requires an intact SigB-system . Although strains with deletions in the agr and/or sarA-system were able to persist at high bacterial numbers in cell culture systems that lack most components of a functioning immune system , they showed severe disadvantages in the in vivo model , as they were unable to defend themselves from invading immune cells ( Fig 2A and 2B ) and were rapidly cleared from the infection focus ( Fig 6A and 6B ) . Consequently , SigB represents a crucial factor to dynamically adapt fully virulent wild-type strains to switch to long-term persistent phenotypes . SigB was described to turn down the agr system [36] , which is most likely responsible for the enhanced inflammatory activity . The downregulated agr-system helps the bacteria to form biofilm [62] and silences aggressiveness for persistence within the host cell [63] . This expression pattern ( high sigB and low agr ) of global regulators appears to be characteristic for SCVs and seems to represent a general adaptation response , as it had been described for stable SCVs generated by aminoglycoside treatment as well [35] . Yet , the varying stress conditions encountered by S . aureus on its way to the intracellular location are less defined and many questions remain to be answered to fully elucidate the complete dynamic bacterial adaption strategies: e . g . , ( i ) which intracellular conditions affect the staphylococcal regulatory factors ? ( ii ) which staphylococcal regulatory factor ( s ) is/are directly influenced by the intracellular milieu ( agr or sigB or further systems , such as the mazEF toxin-antitoxin module [64] ) ? ( iii ) How are the changes of regulatory factors transferred to an increased SCV formation ? ( iv ) which factors of the SigB regulon are required for persistence [65] ? ( v ) Does SigB increase bacterial resistance against antibiotics ? All questions require extensive additional laboratory work to decipher the bacterial adaptation mechanisms in more detail . In our study we used different bacterial backgrounds and mutants , as well as different in vitro and in vivo infection models to demonstrate that bacteria apply general adaption strategies via the crosstalk of regulatory factors with a central function for SigB . By this means , bacteria can rapidly react to changing environmental conditions and dynamically adjust their virulence factor expression at any time of the infection . As the regulatory network involving SigB appears to be the central factor that enables the bacteria to persist and cause chronic infections , it represents a novel therapeutic target for prevention and treatment of chronic and recurrent infection .
The bacterial strains and mutants used in this study are listed in Supp . S1 Table . All the experiments were performed with wild type and mutants in the background of S . aureus LS1 and S . aureus SH1000 . LS1 is a murine arthritis isolate that has been used in infection models before [66] . The strain SH1000 has a complementation of the rbsU gene in the strain 8325–4 which is deficient in SigB activity ( a stress-induced activity ) due to a mutation in the rsbU gene . This gene encodes for a phosphatase required for the release of the sigma factor SigB from inhibition by its anti-sigma factor RsbW . To create a strain with an intact SigB-dependent stress response , the rsbU gene was restored in S . aureus 8325–4 , with the resulting strain called SH1000 [66] . The antibiotic resistance cassette-tagged agr , rsbUVWsigB , and sarA mutations of RN6911 [67] , IK181 [68] and ALC136 [69] were transduced into LS1 using phages 11 , 80α and 85 , respectively . ΔrsbUVWsigB derivatives were cis-complemented by phage transduction with a resistance cassette-tagged intact sigB operon obtained from strain GP268 [70] using phage 80α . The sarA mutant of SH1000 was constructed by replacing the sarA gene with an erythromycin resistance cassette . The sigB mutant of SH1000 was constructed by transducing the sigB mutation ( sigB::Tn551 ) from RUSA168 into SH1000 . The strains were cultivated in tryptic soy broth ( TSB ) at 37°C with linear shaking at 100 rpm in a water bath ( OLS200 , Grant Instruments , England ) . Strains were grown in two sets . In the first set the wild type and the ∆sigB , ∆agr , ∆sarA and ∆agr/∆sarA mutants were cultured and in a second set the wild type and the ∆sigB∆agr , ∆sigB∆sarA and ∆sigB∆agr∆sarA mutants were cultured . The wild type always served as control . During exponential growth phase bacteria were pelleted by centrifugation and culture supernatants were mixed with 10% final concentration of TCA and proteins precipitated at 4°C overnight . Pelleted proteins were washed five times with 70% ethanol and then incubated for 30 min at 21°C and mixed at 600 rpm in a thermomixer ( Eppendorf , Germany ) . Afterwards , pellets were washed once with 100% ethanol and dried in a speed vacuum centrifuge ( Concentrator 5301 , Eppendorf , Germany ) . Subsequently , protein pellets were dissolved in a suitable volume of 1x UT buffer ( 8 M urea and 2 M thiourea ) and incubated for 1h at 21°C with shaking at 600 rpm in a thermomixer ( Eppendorf , Germany ) . No soluble components were pelleted via centrifugation . Protein concentration was determined according to Bradford [71] . 4 μg of protein were reduced and alkylated with Dithiothreitol and Iodoacetamid prior to digestion with Trypsin . Peptides were purified and desalted using μC18 ZipTip columns and dried in a speed vacuum centrifuge ( Concentrator plus , Eppendorf , Germany ) . Dried peptides were dissolved in LC buffer A ( 2% ACN , in water with 0 . 1% Acetic acid ) and subsequently analysis by mass spectrometry was performed on a Proxeon nLC system ( Proxeon , Denmark ) connected to a LTQ-Orbitrap Velos- mass spectrometer ( ThermoElectron , Germany ) . For LC separation the peptides were enriched on a BioSphere C18 pre-column ( NanoSeparations , Netherlands ) and separated using an Acclaim PepMap 100 C18 column ( Dionex , USA ) . For separation a 86-minute gradient was used with a solvent mixture of buffer A ( 2% Acetonitrile in water with 0 . 1% Acetic acid ) and B ( ACN with 0 . 1% acetic acid ) : 0–2% for 1min , 2–5% for 1 min , 5–25% for 59 min , 25–40% for 10 min , 40–100% for 8 min for buffer B . The peptides were eluted with a flow rate of 300 nL/min . The full scan MS spectra were carried out using a FTMS analyzer with a mass range of m/z 300 to 1700 . Data were acquired in profile mode with positive polarity . The method used allowed sequential isolation of the top 20 most intense ions for fragmentation using collision induced dissociation ( CID ) . A minimum of 1000 counts were activated for 10 ms with an activation of q = 0 . 25 , isolation width 2 Da and a normalized collision energy of 35% . The charge state screening and monoisotopic precursor selection was rejecting +1 and +4 charged ions . Target ions already selected for MS/MS were excluded for next 60 seconds . Analysis of MS data was performed with Rosetta Elucidator version 3 . 3 . 0 . 1 ( Rosetta Biosoftware , MA , USA ) . An experimental definition for differential label free quantification was created with default settings: instrument mass accuracy of 5 ppm , spectral alignment search distance of 4 minutes , peak time width minimum of 0 . 1 . For identification the S . aureus NCTC 8325 FASTA sequence in combination with SEQUEST/ Sorcerer was used . Oxidation of methionine , carbamidomethylation and zero missed cleavages were specified as variable modifications . After identification an automatic Peptide/Protein Tellers annotation was performed and only Peptide Teller results greater than 0 . 8 were used . At least two peptides per protein or one peptide with protein sequence coverage of at least 10% were necessary for reliable protein identification . Protein intensities received with the Rosetta Elucidator software were further processed using Genedata Analyst version 7 . 6 ( Genedata AG , Basel , Switzerland ) . Protein intensities were normalized using the central tendency normalization with a dynamic target and the median as central tendency . For relative quantification the ratio of the protein intensities between the wild type and the respective mutants from set one or two were calculated . A protein with a ratio of ≥2 was assigned to be upregulated in the wild type and with a ratio of ≤0 . 5 was assigned to be upregulated in the mutant strain . For prediction of protein localization the PSORT database was used ( PSORTdb 3 . 0 , http://db . psort . org/browse/genome ? id=9009 ) . Furthermore , for categorizing the proteins into subgroups The SEED ( Overbeek et al . , Nucleic Acids Res 33 ( 17 ) ) annotation for S . aureus NCTC8325 was used . A heat map with log2 ratios of the wild type versus the respective mutants was created using the package heatmap . plus version 1 . 3 in R version 2 . 15 . 1 ( http://www . R-project . org ) . For cell culture experiments , bacteria were grown overnight in 15 ml of brain-heart infusion ( BHI ) with shaking ( 160 rpm ) at 37°C . The following day , bacteria were two times washed with PBS and were adjusted to OD = 1 ( 578nm ) . Bacterial supernatants were prepared as described [21] . Briefly , bacteria were grown in 5 ml of brain-heart-infusion broth ( Merck , Germany ) in a rotator shaker ( 200 rpm ) at 37°C for 17 h and pelleted for 5 min at 3350 g . Supernatants were sterile-filtered through a Millex-GP filter unit ( 0 . 22 μm; Millipore , Bedford , MA ) and added to the cell culture medium in the indicated concentrations . For growth curves , Staphylococcus aureus wild-type LS1 , SH1000 and their respective mutants were grown overnight in 15ml of Mueller Hinton infusion ( MH ) with shaking ( 160rpm ) at 37°C . The following day , 100ml of MH were inoculated with each strain in order to obtain the starting optical density 0 , 05 ( 578nm ) . The growth of each strain was monitored spectrophotometrically ( 578nm ) and by plating on blood agar for counting of CFU every hour during 8h . The growth rate ( μ , growth speed ) and the generation time ( g ) for each strain used in this study were calculated according the followings formulas [72] . µ=LnN−LnN0t−t0g=Ln2/µ μ = growth rate g = generation time N = final population N0 = initial population t = final time t0 = initial time Various types of host cells , namely primary isolated human umbilical venous endothelial cells ( HUVECs ) and primary isolated human osteoblasts were cultivated as outlined before [73–75] and were infected with different S . aureus strains as previously described [14] . Briefly , primary cells were infected with an MOI ( multiplicity of infection ) of 50 . After 3 h cells were washed and lysostaphin ( 20 μg/ml ) was added for 30 min to lyse all extracellular or adherent staphylococci , then fresh culture medium was added to the cells . The washing , the lysostaphin step and medium exchange was repeated daily to remove all extracellular staphylococci , which might have been released from the infected cells . To detect live intracellular bacteria at different time points post infection ( p . i . ) host cells were lysed in 3 ml H2O ( for 25cm2 bottles ) or 20 ml H2O ( for 175cm2 bottles ) . To determine the number of colony forming units ( CFU ) , serial dilutions of the cell lysates were plated on blood agar plates and incubated overnight at 37°C . The colony phenotypes were determined on blood agar plates by a Colony Counter Biocount 5000 ( Biosys , Karben , Germany ) . A549 cells stably expressing the escape marker YFP-CWT in the cytoplasm were generated by transduction with lentiviral particles as described [76] . Transductants were passaged three times and subsequently selected by FACS . The phagosomal escape signal is based on the recruitment of YFP-CWT to the staphylococcal cell wall upon rupture of the phagosomal membrane barrier . Recruitment thereby is mediated by the cell wall targeting domain of the S . simulans protease lysostaphin [77] . Accumulation of the fusion protein YFP-CWT was recorded by fluorescence microscopy . For this the A549 YFP-CWT cells were infected in imaging dishes with coverglass bottom ( MoBiTec , Göttingen , Germany ) with the S . aureus LS1 wild type strain or the corresponding mutants for 1 h , followed by lysostaphin treatment to remove all extracellular bacteria . 2 h post infection the YFP-signal was observed with a Zeiss Axiovert 135 TV microscope ( Carl Zeiss , Jena , Germany ) equipped with a 50 W HBO mercury short-arc lamp and a Zeiss filter set ( excitation BP 450–490 nm , beam splitter FT 510 nm , emission LP 515 nm ) . For quantitative assays a 100×/NA 1 . 3 plan-neofluar objective ( field of view 25mm ) was used . Data were acquired with an AxioCam MRm camera and processed using Zeiss AxioVision software . 10 fields of view were recorded per experiment and phagosomal escape was enumerated for 5 independent experiments per strain . Human polymorphonuclear cells ( PMNs ) were freshly isolated from Na-citrate-treated blood of healthy donors . For neutrophil-isolation , dextran-sedimentation and density gradient centrifugation using Ficoll-Paque Plus ( Amersham Bioscience ) was used according to the manufacturer's instruction . Cell purity was determined by Giemsa staining and was always above 99% . PMNs were suspended to a final density of 1×106 cells/0 . 5 ml in RPMI 1640 culture medium ( PAA Laboratories GmbH ) supplemented with 10% heat-inactivated FCS ( PAA Laboratories GmbH ) and immediately used for the experiments . All incubations were performed at 37°C in humidified air with 5% CO2 . To measure cell death induction 1x106 cells/0 . 5 ml PMNs were cultured at 37°C in a 5% CO2 atmosphere in 24-well plates and bacterial supernatants were added as indicated . After 1 h cell death induction was analyzed as described previously [78 , 79] . For the flow cytometric invasion assay , primary human osteoblasts or endothelial cells were plated at 2x105 cells in 12-well plates the day before the assay . Cells were washed with PBS , then cells were incubated with 1 ml of 1% HSA , 25 mM HEPES ( pH 7 . 4 ) in F-12 medium ( Invitrogen ) . The bacteria were grown in BHI , washed and the bacterial suspension ( OD 1 , 540 nm ) was prepared and was added to cells . After the lysostaphin step the infected cells were incubated for 24 h and cell death assays were performed by measuring the proportion of hypodiploid nuclei as described [8] . To measure cell death during the bacterial persistence , this protocol was performed every two days . Hemolysis analysis was performed as described previously with some modifications [80] . The lysis efficacies of human red blood cells were measured using whole culture supernatants of S . aureus LS1 , SH1000 and their respective mutants . Briefly , S . aureus cells were cultured in BHI for 17 h at 250rpm . Staphylococcal cells were centrifuged and the supernatants were used for measuring hemolytic activity . Supernatants ( 100μL ) were added to 100μl human red blood cells ( previously prepared , see preparation of erythrocytes ) . To determine hemolytic activities , the mixtures of blood and S . aureus were incubated at 37°C for 30 minutes . Supernatants were collected by centrifugation at 1000g for 5 min and optical densities were measured at 570nm in an ELISA plate reader . The strain S . aureus Wood46 was used as a positive control . For measurement of cytokine release , osteoblasts were seeded in 12-well plates and stimulated with live staphylococci as described above . After the lysostaphin step , cells were washed and incubated with new culture medium for 24 h . Conditioned media were centrifuged to remove cells and cellular debris , and samples were frozen at 20°C until levels of cytokines and chemokines were measured . The levels of RANTES [regulated on activation of normal T cell expressed and secreted] was measured using immunoassays from RayBiotech according to the manufacturer`s description . Results are given as absolute values ( ng/ml ) of two independent experiments performed in duplicates . Statistical analysis was performed using the ANOVA test . Outbred Wistar adult rats ( groups of 10–12 rats ) weighing 250–350 g were anesthetized with ketamine/xylazine . The left tibia was exposed and a hole made with a high-speed drill using a 0 . 4 mm diameter bit . Each tibia was injected with a 5 μl suspension containing 1x106 CFU of bacteria suspended in fibrin glue ( Tissucol kit 1 ml , Baxter Argentina-AG Viena , Austria ) . Groups of rats were sacrificed at 4 days or 14 weeks after intratibial challenge by exposure to CO2 . Both left and right tibias were removed and morphometrically assessed by measuring the Osteomyelitic Index ( OI ) as described previously [41] . Both epiphysis were cut off from the infected tibias and 1 cm bone segments involving the infected zone were crushed and homogenized . Homogenates were quantitatively cultured overnight on trypticase soy agar and the number of CFU was determined . The OI and the S . aureus CFU counts from each experimental group ( infected and control ) were compared . For RNA extraction , we used the kit RNeasy Mini kit ( Qiagen ) . The RNA extraction was performed following the manufacture instruction and the suggestions of the protocol described by Garzoni et al . [81] . All the primers used in this study are listed in Supp . S2 Table . Real-time PCR was performed by using the RNA isolated from infected cells , infected tissues or different bacterial isolates . The cDNA was obtained using the kit QuantiTect reverse transcription ( Qiagen ) and iQSYBR Green Supermix ( BIO-RAD ) was used . The reaction mixtures were incubated for 15 min at 95°C followed by 40 cycles of 15 s at 95°C , 30 s at 55°C and 30 s at 72°C using the iCycler from BIO-RAD . PCR efficiencies , melting-curve analysis and expression rates were calculated with the BIO-RAD iQ5 Software . In order to analyze the expression of bacterial and host cell genes , respectively , the primers listed in S2 Table were used [14 , 63] . The expression analysis experiments were performed using the software CFX Manager software which calculates the normalized expression ΔΔCT ( relative quantity of genes of interest is normalized to relative quantity of the reference genes across samples ) . The genes used as housekeeping genes to analyze the chemokine expression were GAPDH and B-actin . As controls we used uninfected host cells . For bacterial factors , we used as housekeeping genes aroE , gyrB and gmk . The results shown in each graph are normalized to that control ( control expression = 1 ) . The isolation of human cells and the infection with clinical strains were approved by the local ethics committee ( Ethik-Kommission der Ärztekammer Westfalen-Lippe und der Medizinischen Fakultät der Westfälischen Wilhelms-Universität Münster ) . For our study , written informed consent was obtained ( Az . 2010-155-f-S ) . Taking of blood samples from humans and animals and cell isolation were conducted with approval of the local ethics committee ( 2008-034-f-S; Ethik-Komission der Ärztekammer Westfalen-Lippe und der Medizinischen Fakultät der Westfälischen Wilhelms-Universität Münster ) . Human blood samples were taken from healthy blood donors , who provided written informed consent for the collection of samples and subsequent neutrophil isolation and analysis . For the local osteomyelitis model care of the rats was in accordance with the guidelines set forth by the National Institutes of Health [82] . The experimental protocol involving rats in the experiments was approved by the Institutional Committee for the Care and Use of Laboratory Animals ( CICUAL ) , School of Medicine , University of Buenos Aires ( resolution CD 2361–11 ) . The relationship between WT and all the different mutants was established by the one way ANOVA test with the Dunnett multi-comparison post-test . Significance was calculated using the GraphPad Prism 6 . 0 software and results were considered significant at P = 0 . 05 .
|
Staphylococcus aureus is a frequent pathogen of severe invasive infections that can develop into chronicity and become extremely difficult to eradicate . Chronic infections have been highly associated with altered bacterial phenotypes , i . e . , the small colony variants ( SCVs ) that dynamically appear after bacterial host cell invasion and are highly adapted for intracellular long-term persistence . In this study , we analyzed the underlying mechanisms of the bacterial switching and adaptation process by investigating the functions of the global S . aureus regulators agr , sarA and SigB . We demonstrate that a tight crosstalk between these factors supports the bacteria at any stage of the infection and that SigB is the crucial factor for bacterial adaptation during long-term persistence . In the acute phase , the bacteria require active agr and sarA systems to induce inflammation and cytotoxicity , and to establish an infection at high bacterial numbers . In the chronic stage of infection , SigB downregulates the aggressive bacterial phenotype and mediates the formation of dynamic SCV-phenotypes . Consequently , we describe SigB as a crucial factor for bacterial adaptation and persistence , which represents a possible target for therapeutic interventions against chronic infections .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Sigma Factor SigB Is Crucial to Mediate Staphylococcus aureus Adaptation during Chronic Infections
|
Genotype imputation methods are now being widely used in the analysis of genome-wide association studies . Most imputation analyses to date have used the HapMap as a reference dataset , but new reference panels ( such as controls genotyped on multiple SNP chips and densely typed samples from the 1 , 000 Genomes Project ) will soon allow a broader range of SNPs to be imputed with higher accuracy , thereby increasing power . We describe a genotype imputation method ( IMPUTE version 2 ) that is designed to address the challenges presented by these new datasets . The main innovation of our approach is a flexible modelling framework that increases accuracy and combines information across multiple reference panels while remaining computationally feasible . We find that IMPUTE v2 attains higher accuracy than other methods when the HapMap provides the sole reference panel , but that the size of the panel constrains the improvements that can be made . We also find that imputation accuracy can be greatly enhanced by expanding the reference panel to contain thousands of chromosomes and that IMPUTE v2 outperforms other methods in this setting at both rare and common SNPs , with overall error rates that are 15%–20% lower than those of the closest competing method . One particularly challenging aspect of next-generation association studies is to integrate information across multiple reference panels genotyped on different sets of SNPs; we show that our approach to this problem has practical advantages over other suggested solutions .
Genome-wide association studies have identified many putative disease susceptibility loci in recent years [1]–[3] . This approach to studying disease has succeeded largely because of improved catalogues of human genetic variation [4] and advances in genotyping technology , but it has also been bolstered by the rise of genotype imputation methods [5]–[8] , which have allowed researchers to tease increasingly subtle signals out of large and complex genetic datasets [9] , [10] . Imputation methods work by combining a reference panel of individuals genotyped at a dense set of polymorphic sites ( usually single-nucleotide polymorphisms , or “SNPs” ) with a study sample collected from a genetically similar population and genotyped at a subset of these sites . Figure 1 shows a schematic example of such a dataset . Imputation methods predict unobserved genotypes in the study sample by using a population genetic model to extrapolate allelic correlations measured in the reference panel . The imputed genotypes expand the set of SNPs that can be tested for association , and this more comprehensive view of the genetic variation in a study can enhance true association signals and facilitate meta-analysis [9] , [10] . To date , most imputation analyses have used reference panels composed of haplotypes from Phase II of the International HapMap Project , together with study samples genotyped on commercial genome-wide SNP arrays . Figure 1 depicts this arrangement , which we call Scenario A . To understand how imputation methods work in this setting , it helps to observe that the SNPs exist in a natural hierarchy , such that they can be partitioned into two disjoint sets: a set T that is typed in both the study sample and the reference panel , and a set U that is untyped in the study sample but typed in the reference panel . Informally , most imputation methods phase the study genotypes at SNPs in T and look for perfect or near matches between the resulting haplotypes and the corresponding partial haplotypes in the reference panel—haplotypes that match at SNPs in T are assumed to also match at SNPs in U . This is the fundamental basis of genotype imputation . Several important points emerge from this description . First , the accuracy with which the study haplotypes are phased at SNPs in T should determine how well they can be matched to haplotypes in the reference panel , which should in turn influence the accuracy of imputation at SNPs in U . Second , accounting for the unknown phase of the SNPs in T can be computationally expensive; if the haplotypes at these SNPs were known , most methods would be able to impute genotypes at SNPs in U more quickly . Third , many existing methods do not use all of the available information to phase the study genotypes at SNPs in T . In principle , a phasing algorithm should be able to “learn” about desirable phasing configurations for a given study individual by pooling information across the reference panel and all other individuals in the study , and the phasing accuracy should increase with the sample size; in standard practice , most imputation methods gain phasing information about each study individual only from the reference panel , and phasing accuracy does not depend on the size of the study sample . ( This description applies to imputation methods based on hidden Markov models , or “HMMs” [6] , [11]; non-HMM methods often discard other kinds of information . ) The BEAGLE imputation model [12] , [13] is one notable exception to this point , and we discuss its alternative modeling strategy in detail in this work . We have developed a new algorithm that seeks to improve imputation accuracy at untyped SNPs by improving phasing accuracy at typed SNPs , building on the points raised above . Most HMM-based imputation methods simultaneously estimate missing genotypes and analytically integrate over the unknown phase of SNPs in T . By contrast , we propose to alternately estimate haplotypes at SNPs in T and impute alleles at SNPs in U , assuming the haplotype guesses are correct . We account for the phasing uncertainty in the data by iterating these steps in a Markov chain Monte Carlo ( MCMC ) framework . Separating the phasing and imputation steps allows us to focus more computational effort on phasing and use more of the available information; the extra computation used in this step is largely balanced by the quick haploid imputation in the step that follows . This approach can improve imputation accuracy in Scenario A , as we show in the Results section , but another major motivation of this work is to extend IMPUTE [6] to handle “next-generation” association datasets . By this , we refer to studies in the near future that will have access to additional reference data that could inform imputation . Next-generation reference panels will present new challenges for imputation , including larger sample sizes; unphased and incomplete genotypes; and multiple reference panels containing different SNP sets . Our method aims to use the principles outlined above to address these challenges and improve imputation accuracy in next-generation studies . One new data configuration , which we call Scenario B and explore in detail in the current study , is presented in Figure 2; we will address other next-generation reference panels in the Discussion . In Scenario B , there are different amounts of genotype data in different cohorts of a study . For example , the Wellcome Trust Case Control Consortium ( WTCCC ) is currently performing an association study in which 6 , 000 controls will be genotyped on both the Affymetrix 6 . 0 and Illumina 1 M SNP chips , whereas disease cohorts will be genotyped only on either the Affymetrix 6 . 0 chip or the Illumina 670 k chip . In other words , a large set of controls will be genotyped at a subset of HapMap SNPs , and each case cohort will be genotyped at a subset of the SNPs typed in the controls . Published studies have already employed this design [14] , and it may become more prevalent in the future as common sets of population controls become more widely available . In Scenario B , the study individuals genotyped on a larger number of SNPs can be used as an unphased , or “diploid” , reference panel for imputation in the remaining samples ( which do not necessarily have to be cases ) . As before , we approach such a dataset by partitioning the SNPs into disjoint sets , named with reference to the study sample: a set U1 that is untyped in the study sample and typed only in the haploid reference panel , a set U2 that is untyped in the study sample and typed in both the haploid and diploid reference panels , and a set T that is typed in all samples . We apply the same inference principles to Scenario B as to Scenario A: at each MCMC iteration we phase all of the observed data , pooling information across samples typed on common sets of SNPs to estimate each haplotype pair , then perform haploid imputation assuming that all of the haplotype guesses are correct . One novelty of this scenario is that , at SNPs in U2 , the reference panel may contain thousands of chromosomes , in contrast to HapMap Phase II panels that contain only 120–180 chromosomes each . In principle , this added depth should improve imputation accuracy at SNPs in U2 , with notable gains at rare SNPs . The latter point is especially relevant because rare SNPs are an important source of power in imputation analyses [5] , [6] . Scenario B also introduces the problem of multiple reference panels genotyped on different , hierarchical sets of SNPs . Many next-generation imputation datasets will follow this paradigm , which presents modeling challenges that remain largely unexplored . In the sections that follow , we describe the details of our new method as applied to the scenarios in Figure 1 and Figure 2 . We then compare the method with other imputation approaches on real datasets from the United Kingdom that emulate Scenarios A and B . We show that our method can attain higher accuracy than existing methods in Scenario A , but that the absolute gains are small , which we attribute to the inherent limitations of a small set of reference haplotypes . In an example of Scenario B , we demonstrate that our method can use a large unphased reference panel to achieve higher accuracy than imputation based on the HapMap alone . We also show that our method can impute genotypes more accurately than other sophisticated [11] , [13] and simpler [15] methods applied to the same dataset , and that our approach has higher sensitivity and specificity to detect copies of the minor allele at rare SNPs . In addition , we present results that highlight important practical advantages of our imputation modeling strategy over the one used by BEAGLE . We have implemented our new imputation method as an update to our existing software package IMPUTE; the new program is called “IMPUTE version 2” ( IMPUTE v2 ) . We refer to our previously published method [6] as “IMPUTE version 1” ( IMPUTE v1 ) .
IMPUTE v1 and IMPUTE v2 are freely available for academic use from the website http://www . stats . ox . ac . uk/~marchini/software/gwas/gwas . html In Scenario A , IMPUTE v2 estimates marginal posterior probabilities of missing genotypes by alternately phasing all of the SNPs in T in the study sample ( simultaneously imputing any sporadically missing genotypes ) and then imputing study genotypes at the SNPs in U , conditional on the haplotype guesses from the first step . To explain this process in more detail , we begin by defining , the set of known reference haplotypes at SNPs in T and U ( i . e . , the entire reference panel ) ; , the set of known reference haplotypes at SNPs in T; and , the set of unobserved study haplotypes at SNPs in T . If there are NS individuals in the study sample , their haplotypes at SNPs in T can be represented as , where is the haplotype pair for study individual i . The method begins by choosing initial guesses for the haplotypes in – by default , we choose haplotypes that are consistent with the observed genotype data but phased at random . We then perform a number of MCMC iterations . Each iteration updates every study individual i ( in some arbitrary order ) in two steps: We typically run the method for a relatively small number of burn-in iterations that invoke only the phasing step , followed by a larger number of main iterations that include both steps and contribute to the final imputation probabilities . We investigate the convergence properties of the method in Text S1 , Figure S1 , and Table S1 . In Step 1 , the algorithm phases individual i's observed genotype by sampling from . The model we use to specify this conditional distribution is essentially the same one used by IMPUTE v1 [6] – i . e . , we use a hidden Markov model that is based on an approximation to the coalescent-with-recombination process [16] . This model views newly sampled haplotypes as “imperfect mosaics” of haplotypes that have already been observed . As with IMPUTE v1 , we use an estimated fine-scale recombination map [17] for SNP-to-SNP transition probabilities and a result from population genetics theory [6] for emission probabilities , which model historical mutation . One difference between versions is that IMPUTE v1 analytically integrates over the unknown phase of the genotypes in the study sample , whereas IMPUTE v2 uses Step 1 to integrate over the space of phase reconstructions via Monte Carlo . This step is accomplished for each individual by sampling a pair of paths through the hidden states ( haplotypes ) of the model , then probabilistically sampling a pair of haplotypes that is consistent with the observed multilocus genotype . Path sampling is a standard operation for HMMs , although in this case the calculation burden can be reduced by careful inspection of the equations for the HMM forward algorithm [11] . By default , the state space of the model in Step 1 includes all of the known haplotypes in and the current-guess haplotypes in . The computational burden of these calculations ( both in terms of running time and memory usage ) grows quadratically with the number of haplotypes and linearly with the number of SNPs . We later propose approximations to make these calculations more tractable on large datasets . In Step 2 , the algorithm uses each of the haplotypes in ( which were sampled in Step 1 ) to impute new genotypes for SNPs in U . The HMM state space for this step includes only the reference panel haplotypes . The imputation is accomplished by running the forward-backward algorithm for HMMs independently on each haplotype in and then analytically determining the marginal posterior probabilities of the missing alleles – this process is simply a haploid analogue of the one used by IMPUTE v1 . If we assume that both haplotypes were sampled from a population that conforms to Hardy-Weinberg Equilibrium ( HWE ) , it is straightforward to convert these allelic probabilities to genotypic probabilities for individual i . Across iterations , we can then sum the posterior probabilities for each missing genotype as if they were weighted counts; at the end of a run , the final Monte Carlo posterior probabilities can be calculated by renormalizing these sums . By contrast with Step 1 , the computational burden of these calculations grows only linearly with the number of haplotypes . Consequently , Step 2 can usually avoid the approximations needed to make Step 1 feasible , thereby allowing us to make full use of even very large reference panels . By using both the reference panel and the study sample to inform phasing updates in Step 1 , IMPUTE v2 uses more of the information in the data than most comparable methods [6] , [11] , which typically account for phase uncertainty using only the reference panel . At the same time , each iteration is relatively fast because untyped SNPs are imputed in a haploid framework rather than the more computationally intensive diploid framework that is used by other HMM methods . For example , one iteration of IMPUTE v2 will typically finish faster and use less computer memory than a run of IMPUTE v1 on the same dataset , although IMPUTE v2 tends to be slower than IMPUTE v1 on the whole since the new method requires multiple iterations . We explore the computational burden of the method in detail in the Results section . The structure of the dataset is more complex in this scenario than in the previous one , but we follow the same basic principles of imputation: phase the observed data , then impute alleles in each haplotype separately , conditioning on as much observed data as possible . Here , the goal of the phasing step is to end up with three sets of haplotypes: , the known haploid reference panel haplotypes at SNPs in T , U1 , and U2; , the unobserved diploid reference panel haplotypes at SNPs in T and U2; and , the set of unobserved study haplotypes at SNPs in T . If there are NDR individuals in the diploid reference panel , their haplotypes can be represented as , where is the haplotype pair for diploid reference individual i . The method begins by choosing initial guesses for the haplotypes in and – as before , we choose haplotypes that are consistent with the observed genotype data but phased at random . Each MCMC iteration now includes five steps . First , we update every diploid reference individual i: As is Scenario A , burn-in iterations are used only for phasing ( Steps 1 and 3 ) , while subsequent iterations cycle through all five steps . In this algorithm , each study individual gains phasing information from all other individuals in the dataset , which can lead to very accurate haplotype estimates at typed SNPs when the total sample size is large . Once a study individual has sampled a new pair of haplotypes , the imputation step is broken into two parts: SNPs in U2 are imputed using information from both the haploid and diploid reference panels ( Step 4 ) , and SNPs in U1 are imputed using only the haploid reference panel ( Step 5 ) . This modeling choice highlights a core principle of our inference framework: we allow the method to naturally adapt to the amount of information in the data by conditioning only on observed genotypes , not imputed ones , at each step . As noted above , the HMM calculations underpinning our method require more running time and computer memory as more haplotypes are added to the state space of the model . This can be a problem for the phasing updates , whose computational burden increases quadratically with the number of haplotypes included in the calculation . One solution , implemented in the phasing routine of the MACH software , is to use only a random subset of the available haplotypes for each update . For example , when sampling a new haplotype pair from in Step 1 of our algorithm for Scenario A , we could use a random subset of k haplotypes drawn from to build the conditional distribution , rather than the default approach of using all of the haplotypes . This approximation to the model will generally decrease accuracy , but it will also cause the computational burden of the phasing updates to increase linearly ( for fixed k ) , rather than quadratically , with the number of chromosomes in the dataset . We have developed another approximation that also constrains phasing updates to condition on a subset of k haplotypes . Rather than selecting haplotypes at random , our approach seeks to identify the k haplotypes that are in some sense “closest” to the haplotypes of the individual being updated . In genealogical terms , this amounts to focusing attention on the parts of the underlying tree where that individual's haplotypes are located . The idea is that haplotypes that reside nearby in the genealogical tree will the most informative about the haplotypes of interest . The structure of the underlying genealogical tree is usually unknown ( indeed , knowing the tree would essentially solve the phasing problem ) , so we frame the list of the k closest haplotypes as a random variable that gets updated for each individual at each MCMC iteration . To sample a new phase configuration for diploid individual i , we choose k conditioning states as follows: for each available non-self haplotype ( including current-guess haplotypes for other diploid individuals ) , we calculate the Hamming distance to each of individual i's current-guess haplotypes and store the minimum of these two distances . Then , we use the k haplotypes with the smallest distances to build the HMM and sample a new pair of haplotypes for individual i . The transition and emission probabilities of our model [6] depend explicitly on k . The intuition is that , as k gets larger , jumps between different copied haplotypes should become less likely and those haplotypes should be copied with higher fidelity; this is because a chromosome will coalesce faster into a larger genealogy , leaving less time for recombination and mutation events to occur [18] . The underlying theory assumes that the haplotypes in question were sampled randomly from a population , which is clearly not the case when we select k haplotypes in the manner described above . To account for the fact that these haplotypes will find common ancestors ( going backwards in genealogical time ) more quickly than would k haplotypes chosen at random , we replace k with the total number of available haplotypes when specifying the HMM parameters for a phasing update . We refer to this approximation as informed selection of conditioning states . While this method is built upon genealogical intuitions , we emphasize that no explicit genealogies are constructed in our inference scheme . One way of understanding our approach is by comparison to the phasing method of Kong et al . [19] . Their method uses rule-based techniques to phase putative “unrelateds” by identifying long stretches of identity-by-state ( IBS ) sharing between individuals , under the assumption that such sharing is caused by recent common descent . Our Hamming distance metric can be viewed as a way of identifying near-IBS sharing , and our method combines information across multiple closely related individuals in a model-based way rather than seeking perfect IBS matching between specific individuals . In this sense , our approximation can be viewed as a flexible middle ground between full conditional modeling ( which uses all of the available haplotypes to phase an individual ) and the Kong et al . method ( which may use only a small fraction of the available haplotypes to phase an individual ) . In our experience , imputation based on this informed method for choosing conditioning states is only trivially slower than otherwise identical analyses based on random state selection , effectively because the common HMM calculations take much longer than calculating all pairwise Hamming distances in the informed method . At the same time , the informed method can generally achieve the same phasing accuracy as the random method using many fewer states , or higher accuracy for a fixed number of states ( data not shown ) . This is a major advantage because it is computationally expensive to add states to the model ( i . e . , to increase k ) . We therefore focus on the informed state selection method in this study , with the random method used only during MCMC burn-in , although both approaches are implemented in our software . We conduct an exploration of the parameter settings under informed selection , including the dependence of imputation accuracy on k , in Text S1 , where we also discuss potential limitations of the informed state selection scheme . In order to understand the modeling choices underlying our new imputation algorithm , it is crucial to consider the statistical issues that arise in imputation datasets . For simplicity , we will discuss these issues in the context of Scenario A , although we will also extend them to Scenario B in the Results section . Fundamentally , imputation is very similar to phasing , so it is no surprise that most imputation algorithms are based on population genetic models that were originally used in phasing methods . The most important distinction between phasing and imputation datasets is that the latter include large proportions of systematically missing genotypes . Large amounts of missing data greatly increase the space of possible outcomes , and most phasing algorithms are not able to explore this space efficiently enough to be useful for inference in large studies . A standard way to overcome this problem with HMMs [6] , [11] is to make the approximation that , conditional on the reference panel , each study individual's multilocus genotype is independent of the genotypes for the rest of the study sample . This transforms the inference problem into a separate imputation step for each study individual , with each step involving only a small proportion of missing data since the reference panel is assumed to be missing few , if any , genotypes . In motivating our new imputation methodology , we pointed out that modeling the study individuals independently , rather than jointly , sacrifices phasing accuracy at typed SNPs; this led us to propose a hybrid approach that models the study haplotypes jointly at typed SNPs but independently at untyped SNPs . We made the latter choice partly to improve efficiency – it is fast to impute untyped alleles independently for different haplotypes , which allows us to use all of the information in large reference panels – but also because of the intuition that there is little to be gained from jointly modeling the study sample at untyped SNPs . By contrast , the recently published BEAGLE [13] imputation approach fits a full joint model to all individuals at all SNPs . To overcome the difficulties caused by the large space of possible genotype configurations , BEAGLE initializes its model using a few ad-hoc burn-in iterations in which genotype imputation is driven primarily by the reference panel . The intuition is that this burn-in period will help the model reach a plausible part of parameter space , which can be used as a starting point for fitting a full joint model . This alternative modeling strategy raises the question of whether , and to what extent , it is advantageous to model the study sample jointly at untyped SNPs . One argument [20] holds that there is no point in jointly modeling such SNPs because all of the linkage disequilibrium information needed to impute them is contained in the reference panel . A counterargument is that , as with any statistical missing data problem , the “correct” inference approach is to create a joint model of all observed and missing data . We have found that a full joint model may indeed improve accuracy on small , contrived imputation datasets ( data not shown ) , and this leads us to believe that joint modeling could theoretically increase accuracy in more realistic datasets . However , a more salient question is whether there is any useful information to be gained from jointly modeling untyped SNPs , and whether this information can be obtained with a reasonable amount of computational effort . Most imputation methods , including our new algorithm , implicitly assume that such information is not worth pursuing , whereas BEAGLE assumes that it is . We explore this question further in the sections that follow .
As an example of Scenario A , we used the 120 HapMap CEU parental haplotypes as a reference panel to impute genotypes in the WTCCC 1958 Birth Cohort ( 58 C ) controls [1] . The 58 C samples were genotyped on the Affymetrix 500 K SNP chip , and the data were subjected to the SNP and sample filters specified in the WTCCC study [1] . Of the 1 , 502 58 C individuals , 1 , 407 were also genotyped on the Illumina 550 K chip , and 1 , 377 passed filtering in both datasets . We supplied only the latter set of individuals to the imputation methods , and we asked them to impute the 22 , 270 CEU HapMap SNPs on chromosome 10 that were represented on the Illumina chip but not the Affymetrix chip . We then used the imputed Illumina genotypes to evaluate the success of imputation based on the Affymetrix data . We simulated Scenario B by modifying the WTCCC 58 C dataset as follows: First , we integrated the genotypes from the two SNP chips for the 1 , 377 shared 58 C individuals ( see Text S1 for details ) , yielding a consensus set of 44 , 875 SNPs . Next , we split the 58 C samples into two groups: a diploid reference panel of 918 individuals ( 2/3 of the dataset ) and a study sample of 459 individuals . To complete the reference panel , we added 120 haplotypes from the HapMap Phase II CEU data . We then created two Scenario B study sample datasets by masking the genotypes of SNPs unique to each chip in turn; there were 18 , 489 such SNPs on the Affymetrix chip and 22 , 219 such SNPs on the Illumina chip .
The observation that IMPUTE v2 can achieve lower error rates than IMPUTE v1 in Scenario A validates our new approach . At the same time , the absolute improvement is small , as can be seen in Figure 3 by comparing the separation between IMPUTE v1 and v2 with the separation between IMPUTE v1 and MACH , which typically yield very similar results in our experience . We have also performed separate experiments in which IMPUTE v2 achieves much higher phasing accuracy than IMPUTE v1 at SNPs in T , but where the improvements in HapMap-based imputation of SNPs in U remain modest ( data not shown ) . We suggest that this disconnect between phasing accuracy and imputation accuracy is caused by the inherent limitations of a small reference panel; in other words , we posit that existing models would not attain substantially lower imputation error rates with the current HapMap panel even if we knew the phase of the study genotypes perfectly . In the wake of these results , we suspect that the accuracy improvement of IMPUTE v2 over IMPUTE v1 is not practically meaningful for imputation based on the HapMap Phase II data . However , given that IMPUTE v1's computational requirements scale quadratically with the number of chromosomes in the reference panel while IMPUTE v2's requirements grow linearly , the newer version may become more computationally favorable as baseline reference panels grow in the future . For example , expanding the HapMap reference panel in this study to 800 chromosomes ( which is roughly the size anticipated for each panel in the 1 , 000 Genomes Project ) would lead to similar running times for both versions of IMPUTE , but version 2 would need only 2% of the computer memory required by version 1 . At the same time , IMPUTE v2 would probably achieve higher accuracy , and its computational advantages over IMPUTE v1 would continue to grow with larger reference panels . In our Scenario B dataset , we demonstrated that an expanded reference panel containing thousands of chromosomes can greatly improve accuracy over what is possible based on the HapMap alone , although these gains are limited to the subset of HapMap SNPs that are included on multiple genotyping chips . This finding is consistent with the conclusions of the recent BEAGLE paper [13] . IMPUTE v2 was consistently among the most accurate methods we considered . For example , IMPUTE v2 attained best-guess error rates that were 15–20% lower than those of its closest competitor ( BEAGLE ) in a realistic representation of Scenario B . Rare SNPs are of particular interest because of an increasing awareness that such SNPs may underlie common , complex diseases , and because imputation methods gain the most power over tagging approaches at such SNPs [6] , [11] . Expanded reference panels ought to allow rare SNPs to be imputed much more accurately than they can be with the HapMap panel , and our method is able to exploit this information more effectively than competing methods . Relative to IMPUTE v1 ( which had access to only the HapMap reference panel ) and BEAGLE , the main improvement of IMPUTE v2 is to increase specificity by cutting down on false positive heterozygous calls; relative to fastPHASE and PLINK , the main improvement is to increase sensitivity by cutting down on false negative heterozygous calls . Throughout this study we have touched on the fundamental modeling difficulties that arise in imputation datasets , and we have discussed various strategies that have been proposed to solve these problems . In particular , we have contrasted the BEAGLE approach of full joint modeling with the IMPUTE v2 approach , which phases the observed data jointly but imputes the missing alleles in different haplotypes independently . Based on the results seen here and elsewhere [13] , we claim that BEAGLE gains very little useful information through joint modeling of entire imputation datasets . Consider these lines of evidence: The first two points document strange behavior of the BEAGLE method: apparently , adding data – whether in the form of additional SNPs or additional individuals in the study sample – can cause BEAGLE's imputation accuracy to decrease . More specifically , it seems that increasing the proportion of missing data harms BEAGLE's inferences . This suggests an explanation for the third point above: as the reference panel grew and the study sample remained fixed , the total proportion of missing genotypes in the sample decreased , thereby generating datasets that were relatively less harmful to BEAGLE . In our view , these disparate observations point to a single underlying cause: joint modeling of untyped SNPs is generally ineffective , and it grows progressively worse as the space of missing genotypes expands . BEAGLE was competitive in our analyses , so its modeling strategy may have some merit , but it is also possible that BEAGLE's success came in spite of the joint modeling framework , not because of it . A better alternative might be to embed the same clustering model in a framework like the ones used by fastPHASE or IMPUTE v2 . We suggest that further scrutiny be applied before a full joint model is used in general applications . Comparisons like ours , and others [13] , are necessarily restricted to artificially small datasets , but we have shown that these “toy” datasets can mask problems that might occur in more realistic settings , which will often include larger amounts of missing data . In practice , the accuracy levels and running times achieved by BEAGLE in our study may represent best-case scenarios rather than standard results . These considerations apply to imputation datasets in general , but it is particularly interesting to examine them in the context of multiple reference panels genotyped on different sets of SNPs . BEAGLE's joint approach to such datasets is flexible , but we have seen that it can lose accuracy when certain kinds of new data are added . Conversely , IMPUTE v2's multi-panel modeling strategy responds intuitively to new sources of information like additional individuals or SNPs . This property makes it easy to predict how IMPUTE v2 will perform in larger and more complex datasets than the ones used here , whereas the same cannot necessarily be said for BEAGLE . More broadly , we believe that any imputation algorithm should strive to incorporate as much of the available reference information as possible while remaining easy to use . For example , in Scenario B it is desirable to simultaneously impute the SNPs in the expanded panel ( to improve accuracy ) and the SNPs represented only in the HapMap ( to maintain genomic coverage ) . IMPUTE v2 provides an integrated framework for handling this kind of problem: it is flexible enough to handle numerous variations of Scenarios A and B , yet it remains tractable by focusing computational effort on the parts of the dataset that are most informative . The expanded reference panel we considered was constituted by controls genotyped on multiple SNP chips , but other kinds of new reference panels will also become available in the near future . For example , the HapMap Project has recently augmented its Phase II data with additional samples from both the original HapMap locations and new locations aimed at capturing more human genetic diversity . These samples have all been genotyped on multiple , largely non-overlapping SNP chips , and could be used for imputation in the same way as the controls in our Scenario B . In addition , the 1 , 000 Genomes project is currently pursuing whole-genome sequencing of hundreds of individuals sampled from broad geographic regions in Africa , East Asia , and Europe . One aim of the project is to generate high-quality haplotypes for these individuals , including near-complete coverage of SNPs with population MAFs of 1% or more . This resource will increase the utility of imputation approaches by expanding both the number of chromosomes in the reference set and the number of SNPs that can be imputed . Our method is well-suited to this kind of dataset: in addition to its ability to accurately impute rare SNPs , which will constitute most of the new variants in the 1 , 000 Genomes data , IMPUTE v2 expends relatively little computational effort on haploid imputation steps . This means that , for a given SNP chip typed in a given study sample , doubling the number of untyped variants in a phased reference panel will increase the computational burden of imputation by a factor of less than two . By contrast , other imputation methods ( such as IMPUTE v1 , BEAGLE , and fastPHASE ) would slow down by a factor of at least two . One major use of our new method ( and of imputation methods generally ) will be to facilitate meta-analyses [9] , [10] , which combine samples from studies of similar diseases to increase the chances of detecting low-penetrance risk alleles . For this application , we might expect to repeat Scenario B for a number of different study samples genotyped on different SNP chips . Rather than re-phase the diploid reference panel for each study sample , we can save time by simply storing the posterior samples from a single run of phasing the reference panel , then read these sampled haplotypes from memory when processing each study sample . This functionality is implemented in our software . IMPUTE v2 is already fast enough to use in large association studies , but we also have plans to make it faster . We believe that the software can gain some speed simply by optimizing the code , but we also have plans to implement an analytical speed-up for the HMM forward-backward calculations [22] that may further decrease running times by a factor of five or so . Finally , while we described our imputation approach in terms of two specific scenarios involving the HapMap , it could in fact be generalized to include any number of reference panels of any type ( phased/unphased , complete/incomplete ) so long as their SNP sets follow a hierarchy such as the ones laid out in Figure 1 and Figure 2 . We envision that IMPUTE v2 will be used in a variety of situations . For example , it may soon become standard practice to combine the HapMap Phase II and Phase III datasets to create a compound reference panel like the one in Scenario B , except with all of the reference data phased . Another plausible situation is the version of Scenario B that we described , in which a large set of controls is used to impute genotypes in cases; we discuss some concerns about association testing in this setting in Text S1 and Figure S2 . IMPUTE v2 will also be applied in populations beyond the UK controls used in this study , and we expect that its performance will follow trends much like those observed for similar imputation methods [23] , [24] . Our modeling strategy is flexible and fast , and it is general enough that it could be adopted by other imputation methods . We believe that this intuitive way of thinking about imputation datasets will benefit next-generation association studies , and that IMPUTE v2 will prove to be a useful tool for finding subtle signals of association .
|
Large association studies have proven to be effective tools for identifying parts of the genome that influence disease risk and other heritable traits . So-called “genotype imputation” methods form a cornerstone of modern association studies: by extrapolating genetic correlations from a densely characterized reference panel to a sparsely typed study sample , such methods can estimate unobserved genotypes with high accuracy , thereby increasing the chances of finding true associations . To date , most genome-wide imputation analyses have used reference data from the International HapMap Project . While this strategy has been successful , association studies in the near future will also have access to additional reference information , such as control sets genotyped on multiple SNP chips and dense genome-wide haplotypes from the 1 , 000 Genomes Project . These new reference panels should improve the quality and scope of imputation , but they also present new methodological challenges . We describe a genotype imputation method , IMPUTE version 2 , that is designed to address these challenges in next-generation association studies . We show that our method can use a reference panel containing thousands of chromosomes to attain higher accuracy than is possible with the HapMap alone , and that our approach is more accurate than competing methods on both current and next-generation datasets . We also highlight the modeling issues that arise in imputation datasets .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"genetics",
"and",
"genomics/genomics",
"genetics",
"and",
"genomics/bioinformatics"
] |
2009
|
A Flexible and Accurate Genotype Imputation Method for the Next Generation of Genome-Wide Association Studies
|
Mixed intestinal infections with Entamoeba histolytica , Entamoeba dispar and bacteria with exacerbated manifestations of disease are common in regions where amoebiasis is endemic . However , amoeba–bacteria interactions remain largely unexamined . Trophozoites of E . histolytica and E . dispar were co-cultured with enteropathogenic bacteria strains Escherichia coli ( EPEC ) , Shigella dysenteriae and a commensal Escherichia coli . Amoebae that phagocytosed bacteria were tested for a cytopathic effect on epithelial cell monolayers . Cysteine proteinase activity , adhesion and cell surface concentration of Gal/GalNAc lectin were analyzed in amoebae showing increased virulence . Structural and functional changes and induction of IL-8 expression were determined in epithelial cells before and after exposure to bacteria . Chemotaxis of amoebae and neutrophils to human IL-8 and conditioned culture media from epithelial cells exposed to bacteria was quantified . E . histolytica digested phagocytosed bacteria , although S . dysenteriae retained 70% viability after ingestion . Phagocytosis of pathogenic bacteria augmented the cytopathic effect of E . histolytica and increased expression of Gal/GalNAc lectin on the amoebic surface and increased cysteine proteinase activity . E . dispar remained avirulent . Adhesion of amoebae and damage to cells exposed to bacteria were increased . Additional increases were observed if amoebae had phagocytosed bacteria . Co-culture of epithelial cells with enteropathogenic bacteria disrupted monolayer permeability and induced expression of IL-8 . Media from these co-cultures and human recombinant IL-8 were similarly chemotactic for neutrophils and E . histolytica . Epithelial monolayers exposed to enteropathogenic bacteria become more susceptible to E . histolytica damage . At the same time , phagocytosis of pathogenic bacteria by amoebae further increased epithelial cell damage . The in vitro system presented here provides evidence that the Entamoeba/enteropathogenic bacteria interplay modulates epithelial cell responses to the pathogens . In mixed intestinal infections , where such interactions are possible , they could influence the outcome of disease . The results offer insights to continue research on this phenomenon .
Once trophozoites of Entamoeba histolytica reach the host intestine , they can damage the mucosa epithelial layer and spread through the submucosa and the lamina propia and other tissues . Neutrophils and other cells infiltrate the tissue in the vicinity of amoebic lesions increasing the inflammatory response and tissue damage [1] , [2] . In contrast , Entamoeba dispar , an amoeba that colonizes the human intestine together with E . histolytica and that is morphologically indistinguishable and genetically very similar to the latter , is not invasive and does not produce the clinical manifestations of an E . histolytica intestinal infection [3] , [4] , [5] . Search for expression of genes that could be correlated with the difference in pathogenicity between E . histolytica and E . dispar has mainly revealed higher expression in the former , of molecules involved in lysis of target cells , such as the amoebapore and specific cysteine proteinases [6] , [7] , [8] . Nonetheless , E . histolytica trophozoites can also remain as commensals in the intestinal lumen without causing manifestations of disease [9] , [10] . It has been proposed that E . histolytica pathogenicity could be induced by ingestion of bacteria present in the host intestine . In vitro experiments have shown that after phagocytosis of an E . coli non-pathogenic laboratory strain ( Ec346 ) , trophozoites of E . histolytica increased their virulence together with their adhesive properties to target cells [11] , [12] , [13] . However , the same authors reported that long time cultivation with this bacteria strain rendered the amoebae less virulent [13] . In spite of the possible important role of intestinal bacteria in amoebic behavior in their natural habitat , little has been explored or elucidated about responses triggered by bacteria/amoeba interplay that could be important in the induction of tissue invasion and disease . Pro-inflammatory cytokines released by cultured epithelial and endothelial cells after viral or bacterial infections induce structural and functional alterations in non-infected cells . These alterations lead to increased monolayer permeability and disarray of intercellular junctions and the cortical cytoskeleton which would facilitate passage of pathogens [14] , [15] , [16] , [17] . Experimental infection of intestinal animal models with E . histolytica trophozoites has shown that pro-inflammatory cytokines released by epithelial cells are activated by amoebic cysteine proteinases . It has been proposed that activated cytokines would then recruit neutrophils and other inflammatory cells to the sites of infection , suggesting an important role of the host inflammatory response in tissue damage [17] , [18] , [19] , [20] . In regions where amoebiasis is endemic mixed intestinal infections with E . histolytica/enteropathogenic bacteria are common [21] , [22] , [23] , [24] , [25] , [26] . It is also well established that co-infection with the non-pathogenic E . dispar is prevalent in these regions [9] , [10] , [27] . How amoeba/bacteria interplay in these infections could modify disease manifestations by modulating pathogen virulence and the host response , has not been determined . We approached this problem analyzing the interaction of E . histolytica and E . dispar with two pathogenic enterobacteria strains frequently associated with mixed infections , EPEC and Shigella dysenteriae , isolated from infected individuals . The results were compared with those obtained with amoebae that were not interacted with bacteria and with amoebae that phagocytosed a commensal E . coli strain . The response of epithelial cells to bacteria exposure and the effects of this exposure on cell damage by amoebae were then investigated . E . histolytica/enteropathogenic bacteria interactions induced higher virulence of amoebae . Enteropathogenic bacteria altered the epithelial barrier and induced release of chemoattractant molecules for both neutrophils and E . histolytica and better adhesion of amoebae to epithelial cells with subsequent increases of the cytopathic effect . In vivo , one could hypothesize that such conditions might confer higher susceptibility to pathogen invasion and severe disease manifestations . Our observations about survival and escape of infectious bacteria from amoebae , although preliminary , could be an interesting factor to consider when studying the intestinal environment of mixed infections .
E . histolytica HM1-IMSS trophozoites were cultured in TYI-S-33 medium as indicated [28] after their recovery from hamster liver passage and determination of virulence by production of liver abscesses . E . dispar SAW 760 RR , clone 2 , trophozoites were cultured in axenic medium LY-S-2 as reported [29] . The bacteria utilized corresponded to clinical isolates of the commensal Escherichia coli 086:H18 and the pathogenic EPEC B171-0111: NM , kindly donated by Dr . Teresa Estrada ( CINVESTAV , México ) and Shigella dysenteriae kindly provided by Dr . Celia Alpuche ( Pediatrics Hospital , National Institutes of Health , Mexico ) . MDCK ( NBL-2 ) , dog kidney epithelial cells , passage 72 , were grown to form confluent polarized monolayers as previously reported [30] . Cells were seeded on 24-well culture dishes for interaction with amoebae or bacteria , or on Millicel filters for transepithelial resistance measurements . Monolayers grown on glass cover slips were used for fluorescence microscopy observations . To standardize the conditions to measure phagocytosis of bacteria by amoebae and bacteria viability inside amoebae , bacteria were transfected with vector pd2EGFP ( Clontech Laboratories , Palo Alto , CA ) to express green fluorescent protein , as reported [31] . EGFP-expressing bacteria were co-cultured with amoebae in amoeba/bacteria ratios of 1∶100 for different periods of time . Amoebae were then freed of non-phagocytosed bacteria by extensively rinsing the culture wells with PBS containing 5 mM sodium azide and 50 µM gentamycin . Attached amoebae were gently detached with PBS-gentamycin solution and residual extracellular bacteria removed by centrifugation-resuspension cycles in the same solution . Aliquots of the pelleted amoebae were further checked by fluorescence microscopy for absence of bacteria . Pellets were then fixed with 3 . 7 % formaldehyde , washed with PBS and resuspended in 300 µl of PBS . Intracellular EGFP-fluorescence was determined by flow cytometry . The highest fluorescence values corresponding to the highest number of bacteria ingested by amoebae were registered at 2 and 3 h after the interaction . We chose two and a half hours as the optimal time for phagocytosis . Amoebae that phagocytosed bacteria ( not expressing EGFP ) were utilized for all the following experiments . Cell cultures were freed of bacteria following the techniques applied to fluorescent bacteria and proven to be effective for this purpose . For the viability assays , bacteria expressing EGFP were interacted with amoebae in the above conditions . The capacity of bacteria , recovered from amoebae at different times after been phagocytosed , to form colonies on LB-agar plates was measured by CFU assays . At the indicated times , amoeba cultures were freed of non-phagocyted bacteria as described above . Amoebae were lysed with 0 . 12% Triton X-100 in LB medium and 100 µl of serial dilutions ( up to 10−4 ) of the lysate added to LB agar plates and incubated at 37°C . Colonies formed in each plate after 24 h were quantified . One hundred per cent viability corresponds to the number of bacteria forming colonies 2 . 5 h after been phagocytosed by amoebae . Quantification of colonies formed by bacteria strains not expressing EGFP and recovered from amoebae in the above conditions revealed similar viability to that registered for EGFP-expressing bacteria . E . histolytica trophozoites ( 2×105 ) were co-cultured with bacteria strains not expressing EGFP for 2 . 5 h in amoeba/bacteria ratios of 1∶100 . After removal of non-phagocytosed bacteria , as indicated above , amoebae were deposited on confluent MDCK cell monolayers . After one hour of co-cultivation at 37°C , amoebae were removed by keeping the culture dishes in ice for 5 min and extensive rinsing with ice-cold PBS . The remaining intact epithelial cells were quantified by staining with the methylene blue method [32] . The number of intact cells in MDCK cell monolayers not exposed to amoebae was the control for 0 % cytopathic effect . Inhibition by 100 mM galactose or 250 µM E-64 was measured in amoebae incubated for 30 min before their addition to the epithelial monolayers . E . histolytica trophozoites were labeled with 1 . 0 µl of calcein AM ( Molecular Probes , Eugene OR ) incubating at 37°C for 30 min . After washing with PBS , amoebae were checked for viability with trypan blue and co-cultured with each of the bacteria strains for 2 . 5 h . Non-phagocytosed bacteria were removed as described above and amoebae tested for adhesion to formaldehyde-fixed MDCK confluent monolayers for 20 min , as reported [13] . Adhered amoebae were detached by rinsing with ice-cold PBS , fixed with 3 . 7 % formaldehyde and their fluorescence measured at 517 nm by flow cytometry . Adhesion indexes of calcein-labeled amoebae to MDCK monolayers , previously exposed to bacteria for 4 h , as well as those of amoebae pretreated with 100 mM galactose or the Gal/GalNac lectin polyclonal antibody ( 10 µg/10 , 000 amoebae ) were estimated in the same way . After interaction of bacteria as indicated above , trophozoites were fixed with 2% paraformaldehyde for 20 min and rinsed 3X with PBS . A polyclonal antibody to the Gal/GalNac 170 kDa subunit ( H5 ) , kindly donated by Dr . Barbara Mann , was added at 1∶50 dilution in PBS/2% FBS to a suspension of 2 . 5×105 trophozoites and these incubated at 4°C for 40 min . A secondary antibody ( Goat anti-Rabbit IgG tagged with FITC ) was added to the cells at a dilution of 1∶400 and incubated for 45 min at 4°C . Cells were rinsed with PBS and resuspended in 300 µl of PBS/2% paraformaldehyde . FITC fluorescence was measured in a FACSCalibur flow cytometer at emission peak of 520 nm . Ezymatic activity of cysteine proteinases ( CP ) in E . histolytica lysates and in culture medium was analyzed in control amoebae and amoebae co-cultured for 2 . 5 h with the bacteria strains . After removing non-phagocytosed bacteria as indicated above , trophozoites were cultured for 2 h in culture medium without serum and lysed by freeze-thaw cycles in 50 mM Tris-HCl , pH 7 . 2 , 150 mM NaCl , 1 . 0 mM CaCl2 . Two micrograms of each of the trophozoite lysates and 5 µl of their respective culture medium ( freed of debris by centrifugation at 15 , 000 xg ) were loaded in 1% gelatin , 10% polyacrylamide gels using reported conditions to analyze individual CP activities [32] . The clear areas in the gels revealed cysteine proteinase activity by digestion of the gelatin . Enzymatic activity areas were scanned with the SigmaGel Program in gel from three independent experiments . Lysates and culture media were also separated by electrophoresis in Laemmli's 10% SDS-polyacrylamide gels and silver-stained as controls for protein loading . MDCK cells were plated on polycarbonate filters ( 1 . 2 cm diameter , Millipore Co , Bedford , MA ) previously coated with a solution containing 30 mg/ml of rat Type I collagen . After cells reached confluence , approximately after 48 h in culture , bacteria were added in a ratio of 100∶1 . TER was registered as described [30] , before adding bacteria and at different times of co-culture , previous removal of bacteria and addition of fresh culture medium . FITC-labeled annexin V , Rhodamine-phalloidin and DAPI ( Molecular Probes , Eugene , OR ) were utilized to stain cells and monitor apoptosis , cell morphology and organization of the cytoskeleton in MDCK monolayers exposed to bacteria . For this , cells grown on cover slips were fixed with 3 . 7 % paraformaldehyde and stained by standard fluorescence microscopy methods recommended for these indicators . A monolayer irradiated with UV light was the positive control for apoptosis Chemotaxis was assayed in Transwell chambers as previously reported [33] , loading 150 , 000 calcein-labeled amoebae in the upper chambers resuspended in migration buffer ( 50 mM Tris-HCl pH 7 . 2 , 150 mM NaCl , 1 mM Ca Cl2 , 0 . 01% BSA ) . Amoebae that had not ingested bacteria as well as amoebae that phagocytosed bacteria for 2 . 5 h were tested . The lower chambers contained solutions containing 100 ng/ml in migration buffer of recombinant human cytokine IL-8 ( Preprotech Inc . , Rocky Hill , NJ ) , conditioned culture media obtained from MDCK monolayers co-cultured with each of the bacteria strains in the absence of serum or media from overnight bacteria cultures . Bacteria were removed by centrifugation before adding the culture media to the lower chambers . The number of calcein-labeled amoebae that migrated to the lower chambers was determined by flow cytometry . Culture media from monolayers not exposed to bacteria were used as control . Chemotactic index ( CI ) was calculated considering CI = % chemotactic migration/ % random migration . CI = 1 . 0 corresponds to migration to control media . Purified canine neutrophils were obtained from citrate-treated dog blood , separated in 20 % Ficoll gradients . The layer containing neutrophils was further purified by resuspension/centrifugation cycles in PBS and labeled with calcein . Eighty thousand cells were loaded in the upper chambers and chemotaxis assessed as indicated for amoebae . Total RNA was obtained from control MDCK monolayers , monolayers exposed to E . coli strains , exposed to S . dysenteriae or from monolayers incubated with 10 µM of BAY11-7085 ( inhibitor of NFκB activation , Calbiochem , La Jolla , CA ) , previous exposure to Shigella . RNA was extracted with TRIzol ( Invitrogen , Rockville , MD ) following the specifications of the manufacturer . cDNA was synthesized from 5 µg of DNAse I-treated RNA ( DNA-free™ , Ambion Inc . ) in a reaction mixture containing 5 mM Mg Cl2 , 50 mM KCl , 10 mM Tris-HCl , pH 8 . 3 , 0 . 25 mM of each dNTP , 40U of RNAse inhibitor , 0 . 5 µM of oligo-dT-primers and 50U of Superscript II ( Invitrogen ) . The reactions were allowed to proceed for 45 min at 42°C and inactivated for 5 min at 65°C . Amplification of IL-8 cDNA was done by mixing 1 µl of cDNA with 50 µl of PCR buffer supplemented with 2 . 5 mM MgCl2 , 0 . 5 µM each of sense ( 5′ATGACTTCCAAGCTGGCTG3′ ) and antisense ( 5′TCTGAGTTTTCACAATGTGG3′ ) primers ( designed accordingly to the IL-8 mRNA canine sequence , Canis familiaris ) and 1U of Taq-polymerase ( Invitrogen ) . PCR cycle conditions were 30 s at 94°C , 20 s at 45°C and 1 min at 72°C for 32 cycles . Sense ( 5′ATGGATGATGATATCGCCGC3′ ) and antisense ( 5′TTGGGGTTCAGGGGGGC3′ ) primers were utilized for amplification of the canine β-actin cDNA . The resulting RT-PCR products were analyzed in 1% agarose gels stained with a 2 µg/ml solution of ethidium bromide to monitor the presence of the expected size bands corresponding to IL-8 ( 194 pb ) and β-actin ( 338 bp ) . Data are presented as means±standard deviation . The significance of the results was calculated by t–Student test utilizing the program Sigma Stat . *p values≤0 . 05 were considered significant respect to controls . n values correspond to at least 3 independent experiments done in duplicate . Unless otherwise specified , all reagents were obtained from Sigma Chemical ( St . Louis , MO ) .
After a 2 . 5 h incubation of E . histolytica and E . dispar trophozoites with E . coli 086∶H18 ( Ec ) , EPEC , or Shigella dysenteriae ( Ed ) , the cytopathic effect of amoebae on MDCK cell monolayers was quantified and expressed as percentage of cell damage ( Figure 1 ) . Amoebae that were not incubated with bacteria ( Eh or Ed ) were controls for damage inflicted by amoebae that phagocytosed bacteria . Phagocytosis of E . coli 086∶H18 by E . histolytica ( Eh/Ec ) increased its cytopathic effect , but the increase was not significantly higher than that caused by control amoebae ( 49 . 6±0 . 95 versus 46 . 0±1 . 08 ) . However , after phagocytosis of EPEC ( Eh/EPEC ) or Shigella ( Eh/Sd ) , the cytopathic effect of E . histolytica increased to 64 . 6±2 . 63 and 77 . 6±1 . 24 , respectively . In contrast to what was observed with E . histolytica , phagocytosis of any of the bacterial strains by E . dispar did not induce cytopathic behavior ( Ed/Ec , Ed/EPEC , Ed/Sd ) . Figure 1 also shows that 100 mM Galactose , a known ligand of the amoebic surface Gal/GalNAc lectin , drastically reduced the cytopathic effect of control amoebae to less than 1 . 0 % . The same concentration of the sugar inhibited cell damage 5 % , 13 % and 29 % in amoebae that phagocytosed the commensal E . coli , EPEC or Shigella , respectively . Furthermore , incubation of amoebae with E-64 , a specific inhibitor of cysteine proteinase activity at concentrations not deleterious to amoebic viability [7] , [32] , also caused significant inhibition of the cytopathic effect: 82 % in control amoebae , 76 % , 66 % and 55 % in E . coli , EPEC and Shigella , respect to their cytopathic effect in absence of E-64 . These results showed that phagocytosis of bacteria did not induce virulence in E . dispar , while in E . histolytica it induced an increase of the cytopathic effect particularly after phagocytosis of pathogenic bacteria . Galactose and E-64 inhibition suggest a role for the lectin and cysteine proteinase in the induction of enhanced virulence . As E . dispar did not show induction of virulence in any of the conditions tested , the following results refer only to E . histolytica . Since amoebic adhesion to target cells depends on the activity of the Gal/GalNAc lectin , we analyzed its concentration in E . histolytica trophozoites before and after phagocytosis of bacteria . Amoebae were fixed and labeled with a polyclonal antibody directed to the 170 kDa heavy subunit of the lectin that contains the galactose-binding domain , and a secondary antibody tagged with FITC . Fluorescence intensity on the surface was measured in 10 , 000 cells for each condition . Figure 2A shows the mean index of fluorescence ( MIF ) in a representative experiment . Amoebae that were not exposed to bacteria ( Eh ) showed a MIF = 24 . 7±4 . 61 . Amoebae that phagocytosed E . coli ( Eh/Ec ) showed a MIF = 64 . 42±6 . 0 . Amoebae that phagocytosed EPEC ( Eh/EPEC ) showed a MIF = 73 . 65±5 . 60 and those that phagocytosed Shigella ( Eh/Sd ) showed a MIF = 94 . 42±7 . 34 . The average of MIF values , obtained in three independent experiments , indicated 3 . 0-fold and 3 . 9-fold increases in amoebae after phagocytosis of EPEC or Shigella . Phagocytosis of the commensal E . coli only induced a 2 . 0-fold increase . Treating amoebae with only the secondary antibody did not increased MIF values . Augmented Gal/GalNAc lectin concentration on the surface of amoebae could result in better adhesion and higher cytopathic effect , thereby adhesion of amoebae that phagocytosed bacteria to MDCK cells was quantified ( Figure 2B ) . The adhesion index for control amoebae ( not incubated with bacteria , Eh ) was given value 1 . 00 . After phagocytosis of the commensal E . coli , the adhesion index of trophozoites was 2 . 17±0 . 12 . After phagocytosis of EPEC , the adhesion index was further increased to 2 . 52±0 . 21 and for amoebae that phagocytosed Shigella it reached values of 3 . 65±0 . 18 . The specificity of the adhesion was corroborated in assays where amoebae ( control as well as those that phagocytosed bacteria ) were preincubated with the polyclonal antibody to the amoebic Gal/GalNAc lectin before interaction with the cells . The competition with the antibody reduced adhesion indexes in all the cases to 50% of the control value . An irrelevant antibody of the same isotype did not compete the adhesion of amoebae . These results showed that phagocytosis of bacteria , but particularly pathogenic bacteria , induced a higher concentration of the Gal/GalNAc lectin on the surface of amoebae that seems correlated with a lectin-mediated increase of amoebic adhesion to MDCK cells . However , the antibody could only decrease binding in all the cases to the same level , suggesting that increased adhesion of amoebae after phagocytosis of bacteria is mainly , but not completely lectin-dependent . The results in Figure 1 showing that the increase in cytopathic effect of amoebae that phagocytosed bacteria could be inhibited by E-64 , led us to analyze cysteine proteinase ( CP ) activitiy in these amoebae . Figure 3 shows representative gelatin zymograms of lysates and culture media of E . histolytica trophozoites after phagocytosis of bacteria . The enzymatic activity of each proteinase , corresponding to the area of gelatin digested , was quantified by densitometry ( Figure 3A and 3B ) . The bars in Figure 3C and 3D , express the fold-increase over value 1 . 0 , given to each of the digested areas in lysates and culture media from control amoebae not exposed to bacteria ( Eh ) . Proteinase activity bands corresponding to 48 , 35 , 29 and 27 kDa have been identified in lysates of trophozoites as the major proteinases CP1 , CP2 , and CP5 [6] . Figure 3C shows the values obtained from 3 separate gels where the enzymatic activities of CP1 and CP2 increased above two and three-fold in lysates of amoebae that phagocytosed EPEC or Shigella . The bands of 29 and 27 kDa corresponding to CP5 showed a lower but significant increase above control amoebae that however , was not significant between EPEC and Shigella for the 29 kDa band . After phagocytosis of the commensal E . coli the increase in proteinase activities was not significant . The highest fold-increase for all the activities was observed after phagocytosis of Shigella . The enzymatic activity band of 70 kDa , present in both zymograms , may represent an aggregate of some of the major proteinase activities , as it does not correspond to characerized proteinases . Cysteine proteinases released to the medium of amoebae cultured in axenic conditions have been identified as CP1 , CP3 and CP5 [6] . Zymograms of culture media without serum in which the amoebae were kept for 2 hours after phagocytosis of bacteria ( Figure 3B ) , showed an increase close to 2-fold or higher for all the bands in Shigella . For EPEC the increase was lower , but still significant for bands of 48 kDa , 35 kDa and 27 kDa ( Figure 3D ) . Parallel 10% SDS-polyacrylamide gels of lysates and culture media were silver stained to visualize all the protein bands in the gels ( Figure S2 ) . The gels show that the same protein concentration was loaded in all lanes and no particular difference was observed in particular bands . Therefore , the differences in enzymatic activity in the zymograms are real , indicating specific activation of some proteinases in the amoebae that phagocytosed bacteria . The observation that E-64 substantially inhibited the increase in cytopathic effect induced by phagocytosis of bacteria , shown in Figure 1 , supports that these enzymes could have an important role in the induction of higher virulence . From the results above , it is possible to think that in mixed amoeba/bacteria infections , the interplay of pathogens might modulate damage to the epithelial cells , up-regulating expression of specific pathogenic molecules in the amoebae . However , it is known that epithelial cells exposed to pathogens respond in different ways to their presence . To investigate this , we measured damage of MDCK cells by E . histolytica trophozoites when the monolayers had been previously exposed to the enteropathogenic bacteria used in this study . We analyzed the interaction with amoebae that were not exposed to bacteria as well as with amoebae that had phagocytosed bacteria . As shown in the first set of bars in Figure 4A , damage to control monolayers by amoebae that phagocytosed bacteria was increased ( compare Eh/MDCK with bars in the same set ) corroborating results shown in Figure 1 . The second set of bars shows that amoebae that had not ingested bacteria , but were co-cultured with epithelial cells exposed to bacteria , increased their cytopathic effect ( compare values in this set of bars with those in the left ) . The third set of bars shows that amoebae that had phagocytosed bacteria , when co-cultured with monolayers exposed to pathogenic bacteria , greatly increased cell damage . In this case , there was not only more damage , but it occurred faster , as the monolayers were completely destroyed in 45 min by amoebae that had phagocytosed EPEC or Shigella . The increased damage to monolayers exposed to bacteria suggested that the presence of bacteria could be inducing changes in the epithelial cells that facilitated adhesion of amoebae . Figure 4B shows in the first set of bars , the adhesion of amoebae that had phagocytosed bacteria to control MDCK cells . As also shown in Figure 2 , these amoebae showed higher adhesion than amoebae that had not phagocytosed bacteria . The second set of bars shows that amoebae not incubated with bacteria adhered better if monolayers had been exposed to bacteria , especially Shigella . The third set of bars shows that after exposure of monolayers to bacteria , adhesion of amoebae that had phagocytosed bacteria reached the highest adhesion index , particularly after interaction with Shigella . These results showed that MDCK cell monolayers exposed to bacteria , but particularly to pathogenic bacteria , are better targets for adhesion and damage by amoebae than unexposed monolayers . These two processes can be further enhanced if these monolayers were incubated with amoebae that had phagocytosed bacteria , in particular the pathogenic strains that , as shown above , also increased amoebic virulence . Neutrophils and macrophages are not the only cells attracted by pro-inflammatory cytokines to participate in an inflammatory response of epithelia . E . histolytica trophozoites are also attracted to intestinal epithelial tissue when inflammatory cells are present [18] , [34] . Recent reports have shown that E . histolytica trophozoites migrate in response to human TNFα and IL-1β [35] , [36] , suggesting a possible role for cytokines in the migration of amoebae to sites where bacteria are present and have initiated an inflammatory response . As shown in Figure 5A , amoebae and neutrophils were induced to migrate by culture media from MDCK cells exposed to bacteria . Culture media from MDCK monolayers exposed to E . coli induced a slight but not significant increase of migration of trophozoites and neutrophils , while culture media from EPEC or Shigella-exposed cells induced 2 times and almost 3 times higher migration of amoebae . Neutrophils were particularly attracted to culture medium from cells exposed to Shigella . Lack of response of amoebae and neutrophils to overnight medium of Shigella ruled out that migration could had been induced by bacterial products in the culture media . It has been reported that MDCK cells release IL-8 when subjected to Salmonella typhimurium infection [37] . The presence of IL-8 in the culture media of MDCK cells exposed to Shigella was corroborated by ELISA assays utilizing a monoclonal antibody to canine IL-8 ( clone 258901 , RD Systems Inc , Minneapolis , MN , kindly donated by Dr . A . Castillo , CINVESTAV ) . The results showed concentrations of the cytokine in three different culture media in the range of 180–200 pg/ml . At the same time , it was found that human IL-8 induced migration of both neutrophils and amoebae , providing support to the chemoattractant role of this chemokine when present in culture media . Furthermore , we found that culture media from cells exposed to Shigella in the presence of the inhibitor of IL-8 mRNA expression , BAY11-7085 [38] , reduced migration of both neutrophils and amoebae by 50% . If the inhibitor was added to culture media of cells after their exposure to Shigella , it had not effect on the migration of the amoebae ( data not shown ) . It has been shown that epithelial cells in culture release pro-inflammatory cytokines , such as IL-8 as a defense mechanism when infected by bacteria [39] , [40] . Shigella infection of intestinal cells activates NFκB through a polysaccharide-dependent innate intracellular response leading to the expression of this cytokine [39] , [41] . Other enteropathogenic bacteria also activate release of cytokines [37] , [39] , [40] , [41] . Changes in transcription patterns induced by pathogens could provide a clear indication of the response mechanisms of infected cells . Our data above showed that enterobacteria induced important changes in amoebae and epithelial cells . The changes induced by pathogenic bacteria and especially by Shigella were always more pronounced . Thus , it was clear that the presence of bacteria was affecting the interaction between amoebae and epithelial cells . We analyzed the expression of IL-8 mRNA in cells exposed to E . coli 086∶H18 , EPEC and Shigella by RT-PCR assays using specific primers for MDCK cell IL-8 mRNA designed for this purpose . We thought that it was very interesting that IL-8 was expressed in cells exposed to bacteria as a defense response , but at the same time this chemokine was acting as chemoattractant for the amoebae . Figure 5B shows the results of a representative experiment ( out of three ) where after 4 h of exposure of MDCK cells to E . coli 086∶H18 , EPEC or Shigella , the expression of IL-8mRNA was differentially induced in cells exposed to enteropathogenic bacteria . The highest expression corresponded to cells exposed to Shigella . The figure also shows that the expression of IL-8 mRNA was inhibited 93% in MDCK cells treated with 10 µM BAY11-7085 before exposure to Shigella . These results revealed that exposure of MDCK cells to bacteria , but particularly to invasive bacteria like Shigella , can induce a signaling process that activates NFκB pathways and expression of IL-8 mRNA . We then analyzed if the induction of IL-8 and its release induced structural or functional damage to the bacteria-exposed cells and how this response might be modulated by the presence of another pathogen . Co-culture of pathogenic enterobacteria with epithelial cells is reported to induce alterations of epithelial organization [16] , [42] . Reorganization of the actin cortical cytoskeleton is closely associated with altered permeability of epithelia and endothelia elicited by the infection [15] , [37] , [40] , [43] , [44] . Signaling pathways and mechanisms activated by pathogens to disrupt the structural organization of target cells or to induce a cell response are relatively well known with pathogenic bacteria . In contrast , these aspects are only beginning to be explored in the case of pathogenic amoebae [45] . Thus , we determined if exposure of MDCK cells to enterobacteria modified structural and functional features of the MDCK monolayers that could explain increased amoebic damage . Figure 6A , shows that exposure of monolayers to all the bacteria strains caused gradual decrease of transepithelial resistance ( TER ) , leading to higher permeability . The initial steady state TER values of approximately 528±33 ohm . cm2 dropped to 380±20 ohm . cm2 in monolayers exposed for 5 h to the commensal E . coli and to196±33 ohm . cm2 after exposure to EPEC or Shigella . Monolayers not exposed to bacteria maintained the initial steady state TER . Changes in the organization of the actin cytoskeleton , known to regulate opening of the tight junctions [30] , [46] , were assessed by staining cells exposed to bacteria with Rhodamine-phalloidin to visualize polymerized actin . As shown in Figure 6B , actin in control monolayers was forming juxtaposed cortical rings and fine actin filaments on the basolateral side of the cells and microvilli on the apical side , all of them characteristic features of confluent polarized MDCK monolayers ( Figure 6B , a ) . In contrast , exposure to EPEC or Shigella ( Figure 6B , b , c ) induced separation of the cell borders , loss of the cortical actin ring and a striking reorganization of actin into thicker filaments . This rearrangement of actin filaments could be explained by bacteria-initiated disruption of the tight junction components and signaling to activate release of pro-inflammatory cytokines . These changes , not necessarily conducive to cell death ( see Figure 6C , a , b ) initiate the response to the presence of the parasite and its control by the release of pro-inflammatory cytokines .
When trophozoites of Entamoeba histolytica invade the host intestinal mucosa , they can cause inflammatory colitis . However , trophozoites can remain in the colon without causing tissue damage . Phagocytosis of bacteria , regularly present in the colonic flora , has been considered a possible stimulus to induce amoebic invasive behavior . In regions were E . histolytica and E . dispar are endemic , it is common that intestinal infections caused by enteropathogenic bacteria occur simultaneously with the presence of amoebae [21] , [22] , [23] , [24] , [25] , [26] . In these conditions , it is also common to find exacerbated manifestations of the infection . It is then feasible that the interplay between pathogens modulates amoebic virulence and the response of the intestinal epithelial cells . To approach this important aspect of mixed infections , so far poorly examined , we utilized an in vitro system where we could test virulence of E . histolytica and E . dispar trophozoites after phagocytosis of enteropathogenic bacteria and , at the same time , assess if interaction of enteropathogenic bacteria with epithelial cells elicited responses that could modify amoebic damage . The E . coli , commensal 086∶H18 strain , the pathogenic EPEC , as well as an invasive isolate of S . dysenteriae were chosen for this study . Although the natural habitat of EPEC is the not the colon , this non-invasive pathogen was readily phagocytosed by amoebae and it is often present in mixed intestinal infections and has been utilized for in vitro bacterial infection of MDCK monolayers [14] , [30] , [37] , [47] . Additionally , MDCK monolayers are a well characterized epithelial system , known to retain structural , functional and molecular features of polarized epithelia . A surprising first result was that only E . histolytica trophozoites digested the phagocytosed bacteria , although Shigella retained 70% viability for more than 12 h while , all the bacteria phagocytosed by E . dispar were fully viable after 24 h ( Figure S1 , Text S1 ) . Survival mechanisms of Salmonella and Shigella to escape digestion in highly phagocytic cells are well characterized [47] , [48] , [49] . Although a similar situation may prevent digestion of Shigella by E . histolytica , no studies have been done respect to the mechanisms utilized by pathogenic bacteria to survive in Entamoeba . To our knowledge , this is the first report to address this interesting phenomenon , at the moment beyond the scope of this study . In spite of their different digestive capacity , trophozoites of E . histolytica that had phagocytosed EPEC , and particularly Shigella , caused higher damage to MDCK cells . In contrast , E . dispar , under the same conditions , did not cause any noticeable cytopathic effect . E . histolytica trophozoites have on their surface a protein complex with cell adhesive properties , the Gal/GalNAc lectin [13] , [50] . So , increased adhesion to cells could reflect higher levels of the lectin on the trophozoite surface . Indeed , higher levels of the lectin were found in amoebae that showed increased adherence and caused more damage to target cells . Competition experiments showed that although increased adhesion was related to increased lectin expression on the surface , binding was also due to other molecules on the amoebae . Competition of the binding by galactose , the main ligand of the lectin , supported this conclusion . Expression of cysteine proteinases CP2 and CP5 in trophozoites is also a factor in cell damage [7] , [51] , [52] , although with exception of CP5 , other major CPs are also expressed in the non-pathogenic E . dispar [6] . Analysis of cysteine proteinase activities in lysates and culture media of amoebae that phagocytosed bacteria showed a selective increase of some of the major activities , both in amoebic lysates and culture media . Therefore , one possibility could be that the increased cell damage inflicted by amoebae that phagocytosed EPEC and Shigella is due to higher CP activities released during increased adhesion to target cells . Specific inhibition of cysteine proteinases in co-cultures of amoebae and epithelial cells blocked the increase of cytopathic activities shown by amoebae after ingestion of pathogenic bacteria . In contrast , no significant increases in CP activities were found in E . dispar which could not digest phagocytosed bacteria ( data not shown ) . These data are very suggestive of an induction of the activity of these proteins after phagocytosis and digestion of enteropathogenic bacteria . To this moment , analyses of amoebic microarrays and phagosomes have not provided any clues for the presence of molecules that could explain the increase in amoebic virulence and the survival of Shigella in the E . histolytica [47] , [53] . It has been shown that pro-inflammatory cytokines are released by intestinal cells exposed to E . histolytica infection [18] and that recombinant amoebic proteinases are capable of cleaving pro-inflammatory cytokine precursors to their active form in vitro [34] . It is possible then that increased adhesion of amoebae to epithelial cells together with higher release of CP into the medium , induced by phagocytosis of bacteria and particularly by pathogenic bacteria , would lead to higher concentrations of active cytokines at the sites of contact between amoebae and epithelial cells . The presence of activated inflammatory cytokines would then attract neutrophils and other cells to the sites where amoebae concentrate . Our in vitro model allowed testing the above mentioned possibilities . It has been shown that several pathogens increase permeability of epithelial cells . E . coli ETEC and EPEC strains diminish the barrier functions of cultured epithelial monolayers [16] , [17] . Salmonella and Shigella disrupt the intercellular junctions by alteration of the normal distribution of molecules that associate with them and disrupt the organization of the cytoskeleton [37] , [43] . Virus entry into cells also results in this type of cellular disruption [44] . Alteration of epithelial barriers allows penetration of pathogens into the paracellular space and their dissemination into lower cell layers , as well as migration of inflammatory cells to the luminal side [14] , [16] , [17] , [47] , [54] . E . histolytica trophozoites can also produce decrease of TER in cultured epithelial cells [45] . We tested the effect of enterobacteria on monolayer permeability . TER registers indicated a small decrease in monolayers exposed to the commensal E . coli , and a gradual , but more accentuated drop in monolayers exposed to EPEC or Shigella . After 5 h , TER values had decreased to levels indicative of complete opening of the intercellular junctions . Unsealed monolayers allow passage not only of ions and big molecules , but even of neutrophils and other cells of the immune system . Opening of the tight junctions also allows exposure of receptors for these cells and for pathogens [47] . For example , H . pylori and L . monocytogenes have proteins that bind to E-cadherin once the tight junctions of the epithelial cells are opened , so they can enter cells or epithelial layers [14] , [55] . We found that trophozoites increased their adhesion to MDCK cells that had been exposed to bacteria , ie: with their membrane junctions unsealed . Adhesion was even higher if trophozoites had phagocytosed bacteria . Higher adhesion could be related to the higher cell damage by trophozoites observed in epithelial cells exposed to bacteria , in particular to EPEC or Shigella . These results corroborate that an increase of adhesion to cells by amoebae results in more cell damage . The increase was induced in trophozoites after phagocytosis of bacteria or after exposure of epithelial cells to the same bacteria . Moreover , the highest adherence and cell damage were observed when both , trophozoites and epithelial cells were incubated with bacteria . As shown above , phagocytosis of bacteria induced higher levels of the Gal/GalNAc lectin on the trophozoite's surface , which would facilitate adhesion of the amoebae . However , this would require higher number of receptors on the target cells for a better interaction . Our data suggest that other molecules on the amoebic surface could be participating in the interaction . Preliminary results , currently investigated in our laboratory , have shown that amoebae can induce exposure of TLRs on the surface of intestinal epithelial cells . The gradual drop of TER in monolayers exposed to EPEC or Shigella , suggests a gradual effect on the disorganization of the intercellular junctions that was not apoptotic , but capable of inducing a marked reorganization of the actin cytoskeleton . Disruption of the cortical actin circumferential ring , loss of microvilli and rearrangement of the basolateral filaments , have been correlated with the opening of the tight junctions [30] , [46] , normally a reversible process that allows epithelial cells to adapt to conditions in the medium . However , disorganization of intercellular junctions by enteropathogenic bacteria and other pathogens leads to release of pro-inflammatory cytokines [17] , [18] , [19] , [40] . Our data have shown that MDCK epithelial cells co-cultured with enteropathogenic bacteria suffered functional alterations and released , into the medium , molecules capable of activating chemotaxis of amoebae and neutrophils . Both types of cells were specifically attracted to pro-inflammatory cytokine IL-8 . IL-8 is released by epithelial cells during the interaction with pathogens and acts as a chemokine , playing an important role in the migration of neutrophils to sites of infection [37] , [39] , [40] . We found that MDCK cells exposed to bacteria induced expression of IL-8 mRNA and release of this chemokine to the culture media , corroborating previous results by other authors [39] , [40] , [41] , [44] . Induction was low after exposure to the commensal E . coli , but increased markedly after exposure to Shigella . The induction was almost completely inhibited by inactivation of NFκB . Although this transcription factor activates transcription of different genes [39] , the fact that IL-8 mRNA expression could be differentially induced by the exposure of MDCK cells to different bacteria , and was blocked by the inhibitor of NFκB activation , strongly suggest that bacteria , and particularly Shigella , can activate signaling pathways leading to expression of IL-8 mRNA [15] , [39] , [40] , [47] . We showed here , that canine neutrophils and amoeba migrated in a similar way to human IL-8 and to culture media of MDCK cells previously exposed to bacteria . The highest migration was registered for the culture media from MDCK cells exposed to Shigella where the presence of IL-8 was corroborated by ELISA . Moreover , these cells showed the highest induction of IL-8 mRNA . Previous experiments from our group have shown that amoebae also respond to IL-1β [36] and recently , it has been reported that E . histolytica trophozoites respond to human TNFα gradients by chemotactic sliding [35] . The chemotactic effect of inflammatory cytokines on amoebae supports the idea that amoebae can reach sites in the epithelia where an inflammatory response has been started by bacteria . At the same time , amoebae present in the same milieu can increase their virulence by phagocytosis of bacteria and cells that have been altered by the presence of bacteria are more susceptible to adherence and damage by amoebae . It is possible that all of these phenomena contribute to the pathogen's ability to penetrate epithelial layers . What could be the role of E . dispar in this situation ? The avirulence of E . dispar suggests a non-aggressive participation of this amoeba in mixed infections . However , its ubiquitous presence in samples from patients , its inability to digest ingested bacteria ( Text S1 ) and isolated reports of lesions produced by some amoebic isolates , make study of this amoeba species also worth pursuing . A limitation of our in vitro cell system is the fact that we are observing phenomena outside of the intestine . However , with this approach to mixed amoeba/bacteria infections we have obtained results that could not have been monitored in vivo . We have now a better insight into the role played by the participating elements in the organism . A molecular approach to understand better the signaling processes and the molecules involved at different stages of the infection is now feasible . We hope that our findings encourage research on a health problem still prevalent and neglected in developing countries .
|
In amoebiasis , a human disease that is a serious health problem in many developing countries , efforts have been made to identify responsible factors for the tissue damage inflicted by the parasite Entamoeba histolytica . This amoeba lives in the lumen of the colon without causing damage to the intestinal mucosa , but under unknown circumstances becomes invasive , destroying the intestinal tissue . Bacteria in the intestinal flora have been proposed as inducers of higher amoebic virulence , but the causes or mechanisms responsible for the induction are still undetermined . Mixed intestinal infections with Entamoeba histolytica and enteropathogenic bacteria , showing exacerbated manifestations of disease , are common in endemic countries . We implemented an experimental system to study amoebic virulence in the presence of pathogenic bacteria and its consequences on epithelial cells . Results showed that amoebae that ingested enteropathogenic bacteria became more virulent , causing more damage to epithelial cells . Bacteria induced release of inflammatory proteins by the epithelial cells that attracted amoebae , facilitating amoebic contact to the epithelial cells and higher damage . Our results , although a first approach to this complex problem , provide insights into amoebic infections , as interplay with other pathogens apparently influences the intestinal environment , the behavior of cells involved and the manifestations of the disease .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"infectious",
"diseases/bacterial",
"infections",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"infectious",
"diseases/tropical",
"and",
"travel-associated",
"diseases",
"infectious",
"diseases/protozoal",
"infections"
] |
2008
|
The Interplay between Entamoeba and Enteropathogenic Bacteria Modulates Epithelial Cell Damage
|
The global burden of cholera is largely unknown because the majority of cases are not reported . The low reporting can be attributed to limited capacity of epidemiological surveillance and laboratories , as well as social , political , and economic disincentives for reporting . We previously estimated 2 . 8 million cases and 91 , 000 deaths annually due to cholera in 51 endemic countries . A major limitation in our previous estimate was that the endemic and non-endemic countries were defined based on the countries’ reported cholera cases . We overcame the limitation with the use of a spatial modelling technique in defining endemic countries , and accordingly updated the estimates of the global burden of cholera . Countries were classified as cholera endemic , cholera non-endemic , or cholera-free based on whether a spatial regression model predicted an incidence rate over a certain threshold in at least three of five years ( 2008-2012 ) . The at-risk populations were calculated for each country based on the percent of the country without sustainable access to improved sanitation facilities . Incidence rates from population-based published studies were used to calculate the estimated annual number of cases in endemic countries . The number of annual cholera deaths was calculated using inverse variance-weighted average case-fatality rate ( CFRs ) from literature-based CFR estimates . We found that approximately 1 . 3 billion people are at risk for cholera in endemic countries . An estimated 2 . 86 million cholera cases ( uncertainty range: 1 . 3m-4 . 0m ) occur annually in endemic countries . Among these cases , there are an estimated 95 , 000 deaths ( uncertainty range: 21 , 000-143 , 000 ) . The global burden of cholera remains high . Sub-Saharan Africa accounts for the majority of this burden . Our findings can inform programmatic decision-making for cholera control .
Since the early 1800s , pandemics of cholera have affected millions , with the seventh still ongoing since 1961 [1] . Access to safe water and improved sanitation facilities has eliminated cholera transmission of Vibrio cholerae , the causative agent , in high-income countries . However , the bacteria continue to afflict millions of people in less developed countries where improved water and sanitation infrastructure are not widely available . The actual global burden of cholera is largely unknown as the vast majority of cases are not reported . The World Health Organization ( WHO ) maintains a repository of reported cases and deaths , and publishes annual statistics in the Weekly Epidemiological Record ( WER ) . However , the WHO estimates that only 5–10% of the cases occurring annually are officially reported [2] . This low reporting efficiency is due to a combination of factors including limited capacity of epidemiological surveillance systems and laboratories , and social , political and economic disincentives for reporting [3–5] . Previously , the global burden of cholera was estimated at 2 . 8 million cases of cholera annually , with 91 , 000 deaths [6] . A major limitation of the previous estimate was that the endemic and non-endemic countries were defined solely based on the countries’ reported cholera cases . If a country did not report cases but did , in fact , have cases of cholera , the country was classified as cholera-free . Furthermore , the data used for the previous estimate were from 2000–2008 , prior to the re-appearance of cholera in the Americas . An update of the global burden of cholera is needed to assist public health practitioners and policy-makers in cholera control efforts . In 2010 , a large outbreak of cholera affecting the Americas changed the epidemiology of the disease [7] , and the World Health Organization recommended the inclusion of oral cholera vaccines ( OCVs ) as part of an integrated strategy to control cholera [8 , 9] . In 2013 , a stockpile was created to enable the use of OCV in outbreaks and support for the stockpile was provided by Gavi [10] . This study provides an updated global burden of cholera and addresses some of the limitations of the previous global burden study . It differs from the previous study in that we predicted the presence of cholera cases at the country-level on an annual basis using a multivariable spatial regression model to define the status of a country: endemic , non-endemic , and cholera free . The model allowed for prediction of cholera incidence rate based on the water and sanitation conditions of each country as well as the cholera incidence rate in the 1st order neighboring countries .
A systematic search of all publicly available cholera case and fatality data was conducted . The sources of data included reports to the WHO published annually in the Weekly Epidemiology Records , PubMed , the Global Infectious Disease and Epidemiology Network ( GIDEON ) , the Program for Monitoring Emerging Diseases ( ProMED ) , and Google . The primary sources of data were the WHO’s 2008–2012 WER Annual Cholera Global Surveillance Summaries , which aggregate all cholera cases and fatalities reported to the WHO on an annual basis . This was supplemented with data from GIDEON . The data collected by GIDEON use computer macros which regularly scan source lists including Medline , ProMED , WHO , CDC , national Ministry of Health standard publications , and relevant peer-reviewed journals . Systematic searches of ProMED , PubMed , and Google were also performed using the search terms “cholera” and “acute watery diarrhea . ” The use of multiple sources ensured that all available data were captured from WHO reports , Ministries of Health , research bodies or the media . For annual cases , the source with the highest number of reported cases was used as the final case count . Imported cases were excluded from the case count . The source with the highest number of reported deaths was used as the final death count . All counts were aggregated at the country-level with the exception of India , China , and Indonesia , for which counts were aggregated at the sub-national level . These three countries were analyzed at the sub-national level due to their large population and geographic size , the spatially heterogeneous nature of cholera epidemiology , and the availability of cholera reports at the sub-national level . The 14 states in India , 5 provinces in China , and 4 provinces in Indonesia that reported cholera cases during the study period were included in the analysis . The population data of a country was collected from the United Nations Development Program ( UNDP ) World Population Prospects: The 2012 Revision [11] . This source was selected as it had the most current population estimates ( 2010 ) . The data on accessibility to improved water and sanitation facility were collected from UNICEF’s State of the World’s Children Report 2013 , which had the most current estimates ( 2010 ) [12] . Use of improved sanitation facilities ( % ) was defined as the “percentage of the population using any of the following sanitation facilities , not shared with other households: flush or pour-flush latrine connected to a piped sewerage system , septic tank or pit latrine; ventilated improved pit latrine; pit latrine with a slab; covered pit; composting toilet . Access to improved drinking water source was defined as the percentage of the population using any of the following as the main drinking water source: drinking water supply piped into dwelling , plot , yard or neighbor’s yard; public tap or standpipe; tube well or borehole; protected dug well; protected spring; rainwater; bottled water plus one of the previous sources as a secondary source [12] . For countries with missing data , WHO estimates from 2006 [13] were used . Countries were stratified based on the six WHO regions ( see Fig 1 and Table 1 ) and five WHO mortality strata: ( A ) very low child and low adult mortality; ( B ) low child and low adult mortality; ( C ) low child and high adult mortality; ( D ) high child and high adult mortality; and , ( E ) high child and very high adult mortality [14] . All countries in the A stratum with greater than 95% access to improved sanitation facilities were excluded from the model as the probability of cholera endemicity was assumed to be zero , and ability for secondary transmission of imported cases was also assumed to be zero . A spatial regression model was developed to predict whether or not a country had cholera cases for each year from 2008–2012 . The model’s outcome variable was the log-transformed annual incidence rate . The incidence rate was calculated based on the number of cases reported in a particular calendar year per UNDP estimated total population in the same calendar year , multiplied by a constant to account for 10% reporting efficiency [2] . For countries with no reported cases , the incidence rate of 0 cases/100 , 000 population was replaced with 0 . 001 cases/100 , 000 population , which is below the threshold for sustained transmission in a population and one log below the lowest observed incidence rates , so that the figures could be log-transformed and included in the model . All United Nations member states , with the exception of countries which were in the “A” mortality stratum and had greater than or equal to 95% of their population with sustainable access to improved sanitation , were included in the model . The predictor variables included in the model were 1 ) the percent of the population without sustainable access to improved sanitation , and 2 ) the percent of the population without sustainable access to improved drinking water sources , as defined by UNICEF [12] . Both predictor variables were significantly associated with the outcome variable in a multivariate regression model for every year analyzed from 2008–2012 ( p<0 . 05 ) . The data for the model can be found in supporting information ( see S1 Dataset ) . The spatial regression was run for each year in GeoDa , which is an open source software for spatial data analysis . Spatial weights were created with first order Queen contiguity ( contiguity based on shared border or vertices ) . Queen contiguity was selected over Rook contiguity ( contiguity based on shared border only ) because the absolute length of two countries’ shared border was less important than the proximity of the countries . Initially , we ran ordinary least square ( OLS ) regression , and based on the results of the regression diagnostics for spatial dependence we chose spatial regression model . A threshold incidence rate of 0 . 01 cases/100 , 000 population was established based on the closest match to what we expected the results to be in terms of countries’ classification of endemic , non-endemic , or cholera-free . If the model’s predicted incidence rate exceeded the threshold for a given year in a particular country , that country was determined to have cholera cases in that year . Based on the results of the model , countries were determined to be endemic , non-endemic , or cholera-free . Countries were defined as cholera endemic if they had predicted cholera cases in at least three of the five years under study ( 2008–2012 ) . This definition is in line with the WHO Strategic Advisory Group of Experts on Vaccines and Immunization ( SAGE ) definition of a cholera-endemic country [15] . Non-endemic countries were those with predicted cholera cases in one or two years during the five year study period ( 2008–2012 ) . The population at risk was determined using the percentage of the population without access to an improved sanitation facility . Though access to improved sanitation facilities is not the only determinant of cholera risk , this indicator was selected as a proxy measure due to the lack of availability of other reliable data at the country level [6] . For India , China , and Indonesia , sub-national population data were collected from each country’s national statistics division [16–18] . For these three countries , the population at risk was calculated based on the percentage of the population without access to improved sanitation in the states ( India ) or provinces ( China and Indonesia ) which had reported cholera cases in 2008–2012 . Applying an incidence rate to the entire population at risk in these countries would artificially inflate the global burden of cholera; thus , the population at risk was narrowed to the states and provinces which had reported cases of cholera in the 5-year period under study . In these three countries , the national-level estimate for the percentage of the population without access to improved sanitation was used as a proxy for the sub-national estimate , as sub-national-level data on the percentage of the population without access to improved sanitation were not available . The annual number of cholera cases for the endemic countries ( vide supra ) were estimated from the population at risk multiplied by population-based cholera incidence rate in the country . Population-based cholera incidence rates were obtained from the Diseases of the Most Impoverished ( DOMI ) cholera surveillance program in Kolkata , India [19] , Jakarta , Indonesia [20] , and Beira , Mozambique [21] . These incidence rates were used after a literature review performed to find updated population-based incidence rates for cholera found no recent studies . These data were assumed to still be valid because it is unlikely that cholera incidence rates in these areas changed dramatically between 2005 and 2010 . The DOMI figures included both inpatient and outpatient cases of laboratory-confirmed cholera , and were assumed to be representative of country-wide incidence rates in countries in the same WHO mortality stratum [6] . The population-based incidence rate from Beira , Mozambique was applied to the AFR-E countries . Based on previous analysis , it was assumed that the incidence rate in AFR-D countries was half the incidence in AFR-E countries [6] . The Kolkata , India incidence rate was applied to EMR-D and SEAR-D countries . The Jakarta , Indonesia incidence rate was used as the incidence rate for Indonesia ( SEAR-B ) . For Haiti and the Dominican Republic , the average observed incidence rate from 2010–2012 was used for each country respectively . The reported incidence rates in these two countries ( Haiti and the Dominican Republic ) were assumed to be nearly accurate due to the strength of the cholera surveillance system in these countries leading to a high reporting efficiency , and the high reported incidence rates . All other countries were in stratum B . Based on higher proportions of the population with access to improved sanitation and lower reported incidence rates in these two countries , a 0 . 1 case/1 , 000 population at risk incidence rate was applied to countries in the B stratum . The annual number of cholera deaths for the endemic countries were estimated from the number of cholera cases multiplied by the cholera case-fatality rate . As others have previously noted [6] , cholera mortality is not limited to a certain age group , and is high among all patients . Due to the rapid dehydration of the cholera cases , many deaths occur before these patients are able to reach a health facility . Therefore , it was assumed that facility-based case fatality rates ( CFRs ) were underestimating the true population CFRs . Instead of using facility-based CFRs , CFRs were calculated using inverse variance-weighted average CFRs by WHO mortality stratum ( Table 1 ) . Information on the CFR computation was previously discussed by Ali et al . [6] . Average observed CFRs from 2010–2012 were used for Haiti and the Dominican Republic . Although these CFRs are likely to underestimate the true CFRs , these were used because we assumed that the reporting efficiency is relatively high in these countries as a result of their strong cholera surveillance systems , and because the CFR may be lower than it was in the past three years at the height of the epidemic in these two countries .
The diagnostics for spatial dependence from the OLS model showed considerable spatial dependence ( Moran I = 0 . 3241 , p< . 0001 ) , with significant Robust Lagrange Multiplier error ( p< . 0001 ) . However , Robust Lagrange Multiplier Lag did not yield to a value ( p = . 07 ) , which suggested a spatial error regression model is best fit for the data [22] . Therefore , we applied a spatial error model to predict the incidence for each year . Based on a univariate Moran’s I test with 1st order of neighbor spatial weights , the residuals of the predicted log-transformed incidence rates were spatially random ( p = . 22 ) . There are 1 . 3 billion people at risk for cholera in the 69 countries classified as cholera-endemic . Another 99 million persons are at risk in the three countries the model predicted as non-endemic ( i . e . , Bolivia , Pakistan , and Sri Lanka ) . There are twenty-three endemic countries that have over 10 million persons at risk ( see Table 2 for country classification and country-specific population at risk ) . India , Nigeria , China , Ethiopia , and Bangladesh are the countries with the highest number of people at risk for cholera . The population at risk by WHO region and mortality stratum is shown in Table 2 . There are an estimated 2 . 86 million cases of cholera annually in endemic countries . Spatial distribution of the burden of cholera in endemic countries are shown in Fig 2 . Countries with estimates of more than 100 , 000 cases annually include: India , Ethiopia , Nigeria , Haiti , the Democratic Republic of the Congo , Tanzania , Kenya , and Bangladesh ( for country-level estimates , see Table 2 ) . The WHO regions with the highest burden of cases are AFR-E , SEAR-D , which includes India and Bangladesh , and AFR-D ( Table 3 ) . Haiti alone ( AMR-D ) has a greater burden of cholera cases than all endemic countries in the B-level mortality stratum combined . The average incidence rate in endemic countries is 2 . 30 cases/1 , 000 population at risk per year . Although classified as non-endemic , Pakistan , Bolivia and Sri Lanka were estimated to have a cumulative average of 2 , 737 cases reported annually . Cholera resulted in approximately 95 , 000 deaths annually in endemic countries ( Table 2 ) . This translates to approximately 7 . 50 deaths/100 , 000 population at risk per year in endemic countries . Of the countries with more than 1 , 000 deaths due to cholera annually , all are in the African Region except for India ( SEAR ) , Bangladesh ( SEAR ) , Haiti ( AMR ) , and Sudan ( EMR ) . Countries in the AFR-E stratum have a disproportionate burden of cholera deaths ( see Table 2 for country-specific estimated deaths due to cholera ) . Sensitivity analyses were conducted to produce conservative and liberal estimates in order to avoid over- or under-estimating the annual number of cases and deaths in endemic countries as defined above . The sensitivity analyses included modifications to: a ) population at risk; b ) incidence rates; and , c ) case fatality rates . The first sensitivity analysis established a conservative estimate for population at risk using the fraction of the population with sustainable access to improved water instead of the fraction of the population with sustainable access to improved sanitation . When population at risk was calculated using access to improved water , the total number of cholera cases was 1 . 20 million annually , and the total number of deaths due to cholera was 44 , 000 annually . The liberal estimate for population at risk used the entire population of Indonesia , India , and China ( i . e . , all states/provinces ) to calculate the at risk population in these countries . This liberal estimate increased the total estimated number of annual cholera cases and deaths to 3 . 04 million and 104 , 000 , respectively ( Table 4 ) . The second sensitivity analysis applied incidence rates of 50% and 150% of the original incidence rate . When incidence rates were halved ( conservative estimate ) , the total annual numbers of cholera cases and deaths were 1 . 36 million and 47 , 000 , respectively . The liberal estimate of the incidence rate predicted 4 . 01 million cases and 140 , 000 deaths annually . The third sensitivity analysis assumed a 1% CFR ( conservative estimate ) or a 5% CFR ( liberal estimate ) in all endemic countries . In this analysis , the total number of cholera cases remained the same , but the estimated annual number of deaths due to cholera ranged from 21 , 000 under the conservative CFR assumption to , 143 , 000 under the liberal CFR assumption ( see S2 Dataset for country-specific estimates of the sensitivity analyses in supporting information ) .
We estimate that there were 2 . 9 million cases ( uncertainty range: 1 . 3 to 4 . 0 million ) of cholera annually in 69 cholera-endemic countries and 95 , 000 deaths ( uncertainty range: 21 , 000–143 , 000 ) between 2008 and 2012 . The total estimates are similar to the global burden estimates from 2000–2008 [6] , however the distribution is different . Our study showed that Sub-Saharan Africa accounted for 60% and South-East Asia accounted for 29% of the global burden of cholera cases between 2008 and 2012 . Our findings highlight the fact that cholera remains an important public health issue in more than one-third of the countries of world . Precise estimates of country-level cholera burden remain a challenge for a number of reasons . First is the lack of standard reporting of cholera cases and deaths . The average number of cases and deaths reported by the WHO and other sources from 2008–2012 was 331 , 337 and 6 , 335 , respectively , which was only 11 . 6% of the estimated number of cases and 6 . 6% of the estimated number of deaths . Cases and deaths that do not present to health facilities are not included in the reports and this underestimates the true burden of cholera . In a study in Kenya , there were 46% more cholera cases and 200% more deaths identified through active case finding . Due to the rapid onset of dehydration , death may ensue particularly in areas where accessing health care may be limited by distance , lack of transport or cost [23] . The spatial regression model developed for this global burden estimation predicted the presence of cholera in a country based on the observed incidence rate , the proportion of the population with sustainable access to an improved sanitation facility , the proportion of the population with sustainable access to an improved water source , and a spatial weight which takes into consideration the cholera incidence rates of the neighboring countries . This approach allows for burden estimation in countries which may not report any cases of cholera . The total number of countries estimated to be cholera-endemic is 69 . This figure is similar to the total number of countries which reported cholera cases in 2008–2012 , of which , 42 would be classified as endemic and another 22 would be classified as non-endemic based on reported cholera cases . The estimates produced by this study are slightly higher compared to the previous global burden study ( 2 . 9 million vs 2 . 8 million cases and 95 , 000 vs 91 , 000 deaths ) , and closer to the WHO estimates of 3 to 5 million cases and 100 , 000 to 120 , 000 deaths annually [24] . The regression diagnostics suggest that error of the model are spatially correlated . Thus , spatial dependence enters through the errors in our model and not through the systematic component of the model as the case of spatial lag model . Such a model focuses on estimating the parameters for the independent variables of interest in the systematic part of the model , and disregards the possibility that the observed correlation may reflect something meaningful about the data generation process [25] . The residuals of the predicted log-transformed incidence rates from the spatial regression model were spatially random ( p = 0 . 22 ) suggesting that model adequately addressed spatially correlated error in the model . It also indicates that important independent variables were included in the model and the underlying spatial process that may induce spatial autocorrelation in some of the variables was not missing in the model . There were several limitations in this global burden estimation . There were a few countries which would be classified as endemic if classification were based solely on reported cases , but our model predicted them to be non-endemic ( i . e . , Pakistan ) or cholera-free ( i . e . , Malaysia , Myanmar , Iran , Thailand , and Vietnam ) . This misclassification is due in part to the very low observed incidence rates in these countries , and in part due to the higher proportions of the population with access to improved sanitation and safe water sources than would be expected in cholera-endemic countries . This misclassification may mean that the burden of cholera cases in these countries is regionally disparate or disproportionately affects a small subset of the population . Conversely , some countries for which there is reasonable certainty that they are cholera-free were classified as endemic by our model ( i . e . , St . Lucia , Jamaica , and Tajikistan ) . These countries have high proportions of their populations without access to improved water or improved sanitation , and thus highly likely to become endemic should cholera be introduced from outside . Overall , the three countries assumed to be cholera-free which were classified as cholera-endemic added only 107 cases annually to the global burden . Another limitation of this study is the threshold definition used to determine if a country was predicted to have cholera cases in a given year . The threshold used ( 0 . 01 cases/100 , 000 total population ) was selected based on its accuracy in classifying countries known to be cholera-endemic or cholera-free . Due to the spatial heterogeneity of cholera in some of the larger countries , a country-wide threshold may not accurately predict whether the entire country was cholera-free or cholera-endemic in a given year . A further limitation of this study is the inability to elucidate the highest-risk groups within a country due to the lack of accurate age-specific incidence rates and CFRs . Others have recognized that in many countries , children under five are disproportionately at risk for cholera [6] . This should be taken into consideration when decision-makers are planning cholera control programs . In the absence of reliable country-level incidence rates and CFRs , these rates often had to be taken from other countries within the WHO region or mortality stratum . Spatial and temporal heterogeneity of cholera transmission also means that the country-level burden is rarely constant from year to year , especially in countries with moderate incidence rates . To address this limitation , sensitivity analyses were conducted to generate uncertainty ranges for global estimates of cases and deaths . Finally , the size and socio-economic diversity in the populations of India , China , and Indonesia make predicting the burden in these countries challenging yet critical to the total global estimate . For these countries , the states or provinces at risk were considered to be only those with reported cases between 2008 and 2012 . The incidence and case fatality rates were applied only to these states or provinces . For modeling purposes , the sub-national administrative boundaries were used so that the spatial model would not falsely weigh countries which neighbor cholera-free states or provinces . However , the model is not precise enough to estimate sub-national incidence rates , as the access to water and sanitation data are not available at sub-national levels . This updated global burden estimation is important for policy and public health decision-makers in countries which may not know of the cholera burden in their countries due to challenges in surveillance and case reporting . The updated findings are equally important for decision-makers who are involved in building the OCV stockpile , as it provides an estimate of the size of the demand for the vaccine . The findings are also important for those who are responsible for requesting OCV for their country , either from the manufacturer or from the WHO’s stockpile , and for members of the International Coordinating Group , who are responsible for managing the use of the OCV stockpile . More accurate data on country-level and global-level cholera burden will allow decision-makers to more effectively allocate resources to the countries with the greatest need . Our findings show that the cholera burden remains high . Efforts for an integrated approach to cholera control are vital to prevent further spread of the disease and mitigate the resulting mortality and morbidity from this deadly disease .
|
The global burden of cholera is largely unknown because the majority of cases are not reported . The low reporting can be attributed to limited capacity of epidemiological surveillance and laboratories , as well as social , political , and economic disincentives for reporting . We previously estimated 2 . 8 million cases and 91 , 000 deaths annually due to cholera in 51 endemic countries . A major limitation in our previous estimate was that the endemic and non-endemic countries were defined based on the countries’ reported cholera cases . If a country did not report cases even though the country had cholera , the country was classified as cholera free . This time we addressed this limitation by using a spatial modelling technique , which helped us define the cholera-endemic countries based on access to improved water and sanitation in the country as well as cholera incidence in neighboring countries . Our new estimate illustrates 2 . 9 million of cases and 95 , 000 deaths in 69 endemic countries , with the majority of the burden in Sub-Saharan Africa . The sustained high burden of cholera points to the necessity for integrated and improved control efforts , and these findings may help programmatic decision-making for controlling the disease in endemic countries .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Updated Global Burden of Cholera in Endemic Countries
|
Understanding glycan structure and dynamics is central to understanding protein-carbohydrate recognition and its role in protein-protein interactions . Given the difficulties in obtaining the glycan's crystal structure in glycoconjugates due to its flexibility and heterogeneity , computational modeling could play an important role in providing glycosylated protein structure models . To address if glycan structures available in the PDB can be used as templates or fragments for glycan modeling , we present a survey of the N-glycan structures of 35 different sequences in the PDB . Our statistical analysis shows that the N-glycan structures found on homologous glycoproteins are significantly conserved compared to the random background , suggesting that N-glycan chains can be confidently modeled with template glycan structures whose parent glycoproteins share sequence similarity . On the other hand , N-glycan structures found on non-homologous glycoproteins do not show significant global structural similarity . Nonetheless , the internal substructures of these N-glycans , particularly , the substructures that are closer to the protein , show significantly similar structures , suggesting that such substructures can be used as fragments in glycan modeling . Increased interactions with protein might be responsible for the restricted conformational space of N-glycan chains . Our results suggest that structure prediction/modeling of N-glycans of glycoconjugates using structure database could be effective and different modeling approaches would be needed depending on the availability of template structures .
Glycosylation represents one of the most important post-translational modifications [1] , [2] and is ubiquitous in all domains of life . The glycosylation machinery is largely conserved in eukaryotes , and more than 50% of all eukaryotic proteins are expected to be glycosylated [3] , [4] . An oligosaccharide moiety in a glycoprotein , referred to as a glycan , comes in a diversity of sequences and structures and plays critical roles in a vast array of biological processes [1] . The N-glycosylation pathway is the most common pathway in which an oligosaccharide is covalently attached to the side chain of asparagine [2] . In general , such an oligosaccharide appendage masks the protein surface , protecting the glycoprotein from degradation and nonspecific protein-protein interactions ( reviewed in [5]–[7] ) . N-glycosylation also alters the biophysical properties in the vicinity of the glycosylation site and affects the folding rates and the thermal stability of the protein [8] , [9] . Some N-linked oligosaccharides ( N-glycans ) are directly involved in specific molecular recognition events; e . g . , lectins and antibodies can recognize specific N-glycans on viral envelope glycoproteins such as HIV gp120 [10]–[13] . The impact of glycosylation on the structure of the parent protein and vice versa has been of great interest in structural glycobiology [8] , [14]–[17] . At this time , however , an understanding of which glycans are important components in protein function and how to modify these glycans to optimize the protein properties of interest remain an enigma . Therefore , knowledge of the structure and dynamics of N-glycans is central to understanding protein-carbohydrate recognition and its role in protein-protein interactions . An oligosaccharide chain is flexible in solution and has an ensemble of diverse conformations rather than a single well-defined structure [18]–[20] . The inherent flexibility of oligosaccharides often hinders crystallographic structure determination , and there are only a few crystal structures of oligosaccharides longer than 2–3 residues in the Cambridge Structure Database [21] . In contrast , there are many more crystal structures of glycoconjugates in the Protein Data Bank ( PDB ) [22] , suggesting that the presence of the protein may reduce the conformational freedom of oligosaccharides or even favor a certain conformation over others [23] . For example , the N-glycan conformations in the crystal structures of the Fc domain [24]–[30] exhibit remarkable similarity ( Figure S1 in Supporting Information ) , suggesting that the protein's structure around the glycan has an influence on the glycan's conformation . The number of PDB entries containing carbohydrates has been steadily increasing , but obtaining the complete N-glycan structure remains challenging [23] . Mass spectrometric mapping of N-glycosylation sites is becoming common [4] , providing information about glycosylation sites as well as the relative abundance of different glycoforms . In this context , computational modeling of N-glycan structures is an appealing approach to provide glycosylated protein structure models . In particular , a computational approach that can combine known glycoprotein structures and glycosylation information ( i . e . , glycosylation site , primary glycan sequence , and linkage information ) would be very useful in a variety of applications in glycoscience . For successful template-based glycan structure modeling , it is essential to understand the conformational variability of an oligosaccharide chain when it is glycosylated . In addition , the influence of the protein residues around the glycosylation site can provide valuable insight into the design of new computational approaches that are optimized for glycoconjugates . Several structural database surveys have investigated the general features of N-glycosylation in terms of oligosaccharide and protein structures [14] , [23] , [31]–[35] . In these earlier studies , however , the oligosaccharide conformations were analyzed in terms of individual glycosidic torsion angles , making it difficult to recognize the actual structural variability of glycans en bloc . To the best of our knowledge , the conformational variability of N-glycans using the three-dimensional ( 3D ) structures in the PDB has not been studied . In this work , using the PDB crystal structures that contain N-glycans , we examined the conformational variability in various N-glycans . Using Glycan Reader [36] , an automatic sugar recognition algorithm that we developed , all N-linked glycoprotein structures were obtained from the PDB and sorted by their N-glycan sequence . PDB entries with more than 3 Å resolution were excluded and N-glycan sequences with less than 20 PDB entries were also excluded , resulting in 35 N-glycan sequences ( see the full list in Table S1 in Supporting Information ) . Using random background conformations of each N-glycan sequence , the statistical significance of glycan structural similarity was estimated . The N-glycan structures in the PDB show statistically significant similarity when the local structure around the protein is conserved . When the local protein structures are different , overall N-glycan structures are not conserved , but their internal substructures appear to be strongly conserved due to the proximity to the protein . Our results highlight the applicability of template-based approaches used in protein structure prediction to the structure prediction and modeling of N-glycans of glycoproteins . Although the N-glycan sequences examined in this work mostly represent oligomannose-type glycans due to the limited numbers of crystal structures of complex- and hybrid-type glycans , the conclusions might be applicable to other glycoconjugates' glycan sequences .
The structural similarities of the N-glycans are measured by calculating the glycan RMSD after alignment of the oligosaccharide structures using the carbohydrate ring heavy atoms . N-glycan structural similarity including their orientations with respect to the protein is discussed separately below . Figure 2 shows the RMSD distributions of the N-glycan structure pairs in the PDB and random conformation pool . Note that the RMSD is only measured between glycan structures having an identical sequence . The average RMSD of all PDB structural pairs are 1 . 4±0 . 8 Å . The homologous and the non-homologous N-glycan structure pairs have RMSD values of 0 . 9±0 . 8 Å and 1 . 4±0 . 8 Å , respectively . Both the homologous and non-homologous N-glycans showed smaller RMSD values compared to those in the random glycan structure pool whose RMSD is 2 . 4±0 . 8 Å ( Figure 2A ) . Measuring the structural similarity using RMSD is straightforward , but it is not an objective measure when comparing structures of different lengths and sequences due to its length dependence . When the average RMSD values of the N-glycans are plotted against N-glycan length , i . e . , the number of carbohydrate monomers ( Figure 2B ) , a length dependence is observed for the random background and non-homologous glycan pairs , but homologous glycan pairs do not show such a length dependence . The smaller RMSD values of the homologous N-glycan structure pairs compared to the RMSD values of the non-homologous pairs indicate that the homologous N-glycan structures are more conserved than the non-homologous N-glycan structures . Because our dataset contains different lengths of N-glycan sequences with different branching patterns ( Table S1 ) , we converted the RMSD values to their statistical significance ( p-values ) using the random background glycan structures ( see Methods for details ) . By deriving the statistical significance using the random background having the identical N-glycan sequence , the length dependence is effectively removed . The generalized extreme value distribution ( Eq . 1 in Methods ) was used to estimate the statistical significance [37] , and 35 sets of parameters were determined by fitting the generalized extreme value distribution to the original RMSD distribution of the random conformational pool of each glycan sequence ( see the determined parameters in Table S2 and the fitting results in Figure S2 ) . The calculated p-values ( Eq . 2 in Methods ) represent the probability of having randomly chosen two N-glycan structures whose RMSD is smaller than the random background . A list of p-values and the corresponding RMSD values averaged over different sequences are given in Table 1 . Figures 3A and 3B show the cumulative fraction of homologous and non-homologous glycans structure pairs as a function of their p-value . It is clear that about 67% of the homologous N-glycan structure pairs have a statistically significant level ( p<0 . 05 ) of structural similarity , whereas about 36% of non-homologous N-glycan structure pairs have a statistically significant level of structural similarity . A correlation is also found between the sequence similarity of the glycoprotein and the structural similarity of the N-glycan ( Figure S3 ) . Specifically , about 81% and 91% of N-glycan structure pairs have statistically significant structure similarity when the parent proteins have sequence similarity greater than 50% and 60% , respectively . A similar analysis has been carried out independently using the global distance test ( GDT ) score [38] instead of RMSD , and the conclusion remains the same ( Figure S4 ) . Assuming that the proteins with similar sequences have similar surface features around the glycosylation site , such a high level of N-glycan structure similarity strongly indicates that the protein structure around the N-linked oligosaccharide plays an important role in determining the N-glycan structures . Apparently , not all homologous glycans have significant structural similarity . Figure 4A shows an example of two homologous proteins , the Fc domain of IgG ( PDB:2WAH ) in green and the Fc domain of IgE ( PDB:3H9Y ) in orange , which share a sequence similarity of about 50% and have significantly different glycan structures ( RMSD of 2 . 9 Å and p-value of 0 . 6 ) . The structures of these two homologous proteins around the glycosylation site are similar and well aligned . Notably , the structural difference of the N-glycans arises mainly from the terminal residues at the 1–6 branches ( or 1–6 arm ) . The PDB:2WAH IgG-Fc domain is glycosylated with a different glycoform than typical IgG-Fc glycans whose 1–6 arm carbohydrates are tightly packed with the proteins [24]–[30] . This may explain such a different glycan conformation in PDB:2WAH . There are some non-homologous N-glycan structure pairs that have a statistically significant level of structural similarity . Visual inspection of several examples of non-homologous glycoproteins having similar N-glycan conformations shows no apparent similar protein surface features around the N-glycans . Figure 4B shows an example of two non-homologous glycoproteins , beta-galactosidase ( PDB:3OG2 ) in green and the extracellular domain of the nicotinic acetylcholine receptor 1 subunit ( PDB:2QC1 ) in orange , having a significant level of structural similarity of the N-glycan ( RMSD of 0 . 9 Å and p-value of 0 . 009 ) . Nonetheless , the structure alignment of these two N-glycans results in a poor alignment of the parent proteins . The relative orientation of an oligosaccharide chain with respect to the parent protein can be affected by the Asn side chain conformation and the protein conformation in the vicinity of the glycosylation site . To examine N-glycan structural variability with respect to the parent protein , the heavy atoms of the glycosylated Asn residue were used for alignment of each pair , and then the Euclidean distance of the glycan portion was measured without further alignment . Figures 3C and 3D show the cumulative fraction of structure similarity of the homologous and non-homologous glycans aligned with glycosylated Asn residues . Clearly , structural similarity is greatly reduced when the Asn residues are used for the alignment . Given the fact that glycosylation has a bias towards turns and extended regions [32] , it is not surprising that even homologous N-glycans show reduced structural similarity when the Asn residues are used for the alignment . The observations so far indicate that a comparative modeling approach for N-glycan structures would successfully predict the N-glycan structure itself , especially when the homologous N-glycan templates are present in the PDB , but finding the global orientation of the glycan with respect to the protein would remain challenging . Such difficulties can be significantly alleviated when a partial glycan structure is available . In fact , there are large numbers of partial N-glycan structures available in the PDB , probably due to the removal of glycans prior to structural studies , due to crystallization conditions , or due to missing electron density resulting from flexible glycan structures . For example , as of December 2011 , there were 2 , 517 PDB entries and 10 , 769 N-linked glycan chains in the RCSB database; 84% ( 9 , 027 chains ) had partial glycan structures with less than two carbohydrate units and 15% ( 1 , 394 chains ) of such partial structures showed their parent protein sequence similarity less than 50% . Assuming that one can find such partial glycan structures , Figures 3E and 3F show the cumulative structural similarity of the N-glycans when the first two carbohydrate units in the glycan chains are aligned . Both the structural similarities of the homologous and non-homologous N-glycan structures ( especially the former ) significantly increased , suggesting that the conformations of glycosylated Asn residues and the first few carbohydrates of the N-glycan are important in determining the N-glycan orientations . What makes homologous N-glycan structures conserved compared to non-homologous N-glycans or random background ? Possibly , the protein structures around the glycan may provide a steric barrier , thus restricting the conformational freedom of N-glycans nearby . In addition , specific protein-carbohydrate interactions may play an important role in favoring a certain conformation of the oligosaccharides . If local protein structure around the N-glycan is directly correlated with the N-glycan structure similarity , such information provides valuable criteria in N-glycan structure modeling . Figure 5 shows the correlation between the local protein structure around the glycan chain and the N-glycan structure similarity . As expected , most homologous glycoproteins have similar local protein structures around the glycan chain . However , some homologous N-glycan structure pairs adopt significantly different conformations while their local protein structures are similar ( p-RMSD>0 . 05 and p-local<0 . 01 ) . Visual inspection of such structures shows that the structural differences are mainly due to the terminal residues , especially ones in the 1–6 branches , similar to the case in Figure 4A . The increased flexibility of the 1–6 linkage is not surprising because the 1–6 glycosidic linkage contains three rotatable torsional angles ( compared to two for other glycosidic linkages ) , and the flexibility of the 1–6 linkage has been well documented by other experimental , computational , and structural database surveys [34] , [39]–[42] . To examine the flexibility of different regions of N-glycan structures , we have used the GDT chart [38] . Figure 6 shows two example N-glycan sequences and the corresponding GDT charts , where each bar represents an alignment of an N-glycan pair and the bar is colored according to how well a certain region of the sequence can be aligned each other . Clearly , the increased flexibility of terminal residues is apparent and , in particular , the residues in the 1–6 branches are even more flexible . Non-homologous N-glycan structures in the PDB do not show a correlation with local protein structure around the glycan . There could be several factors responsible for this observation , and the accuracy of local protein structure alignment might be one important factor . To compare the similarity of local protein structure , TM-align [43] was used because the algorithm is general and performed well compared to other local structure algorithms available in our internal testing [44] . However , the TM-align algorithm was developed for comparison of global protein structure , and it is possible that the algorithm is insensitive to the structural similarities of the small number of residues around the glycan chain . Thus , further in-depth investigations with robust local structure algorithms are warranted . The lack of correlation between the local protein structure and non-homologous glycan structures suggests that the gapless threading approach to N-glycan modeling would be inapplicable when no homologous templates are present . It was reported that the majority of glycosylation sites are found to be in convex or flat regions of the protein surface [32] . When the N-linked oligosaccharides are situated in such regions , the terminal residues of a long oligosaccharide may not be able to interact with the protein surface residues , and experience a smaller influence of the local protein environment . Thus , local protein structure around glycan chains might have a stronger impact on the first few residues of the glycan chain rather than on the global structure . Internal substructure conservation can be visualized with the two examples in Figure 6 , showing that the flexibility of the carbohydrate residues increases as the residues move away from the protein . In addition , a large increase in flexibility is observed after the 1–6 linkage , which is known to be flexible . If the N-glycan substructure is more conserved , a threading or fragment assembly approach could be useful to model the N-glycan structures . To quantify the conservation of internal substructures , we compared the structural similarity of the N-glycans as a function of glycan chain length from the protein . Figure 7A shows the average RMSD of N-glycan internal substructures containing only the residues within the given residue distance from the Asn residue of the parent protein . The conservation of the internal substructure is apparent up to 3 or 4 residues away from the Asn residue . Note that N-glycan sequences can have branches , and thus , there could be more residues in a substructure within a certain residue distance . For example , in the two examples in Figure 6 , there are in fact 5 sugar residues at a residue distance of 4 from Asn . To avoid the inherent length dependence of RMSD ( i . e . , a smaller substructure has a smaller RMSD ) , RMSD values for the substructures are converted to p-values using the random background . Figure 7B and 7C show the cumulative fraction of the substructure similarity for homologous and non-homologous N-glycans , respectively . About 80% and 60% of the substructure up to a residue distance of 3 ( black curve ) show significant structural similarity for homologous and non-homologous N-glycans , respectively . The substructures are less conserved when residues up to a distance of 4 are included in the substructure ( blue curve ) . As discussed above , due to its flexibility , the 1–6 linkage might contribute to the diversity of the N-glycan substructures more than other glycosidic linkages . Clearly , when structural similarity of substructures up to a residue distance of 4 is compared without residues linked by the 1–6 linkage ( red curve ) , significant structural conservation is observed even for non-homologous N-glycans . This observation implies that the glycan residues closer to the protein surface have more restricted conformational space and conserved structures .
Elucidation of the factors influencing the conformational variability in N-glycans is essential to understand the dynamics of N-glycans and provides valuable insight into modeling and computational studies of the N-linked oligosaccharides . In this work , we have shown that the conformations of homologous N-glycans are restricted compared to the random background . About 67% of the homologous N-glycan pairs and 37% of the non-homologous N-glycan pairs show statistically significant level of structural similarity . Although excluded from the main analysis , more than 90% of highly homologous N-glycan structure pairs ( protein sequence similarity ≥90% ) show very significant structural similarity ( Figure S5 ) . Why do homologous N-glycans have conserved conformations compared to the free oligosaccharides ? First , protein-carbohydrate interactions may restrict the conformational freedom of the N-glycans . In addition , the shape of the local protein structure may also act as a non-specific steric barrier and restrict the N-glycans to adopt certain conformations . Lastly , crystallographic bias in the dataset could also play a role in conformational similarity of homologous N-glycan structures . Our dataset is composed of crystal structures of well-resolved N-glycan structures; hence , flexible N-glycan structures may not be included in our dataset . Despite the biological importance of N-glycans , understanding the structure and dynamics of N-glycans is currently lacking due to the difficulties in crystallization of glycoproteins and other experimental techniques . The high level of structural similarity among the N-glycan structures found on the surface of homologous proteins strongly indicates that the comparative modeling and threading approach used in protein structure prediction [45]–[47] might perform well in glycan structure modeling if appropriate templates are present . Despite the structural similarity of N-glycans on the homologous glycoproteins , the absolute orientation of N-glycan with respect to the glycosylated Asn residue may differ because the glycosylation site are often found on the loop regions of the protein . N-glycan modeling without good template structures appears to be challenging because of less conserved N-glycan structures found for non-homologous proteins . However , a higher level of internal substructure similarity exists even for non-homologous N-glycan pairs up to a residue distance of 4 without the 1–6 linkage . In fact , these carbohydrate structures that lie close to the protein are key determinants of the overall N-glycan orientation . Thus , a fragment assembly approach might perform well even without homologous N-glycans template structures because of this internal substructure conservation .
Extracting structural information of glycans from the PDB is nontrivial due to a lack of standardized nomenclature and the way the data is presented in the PDB . To recognize the PDB entries that contain carbohydrate molecules , we used Glycan Reader for automatic sugar identification [36] . Briefly , in Glycan Reader , the topologies of the molecules in the HETATM section of a PDB file are first generated using the CONECT section of the PDB file , and the candidate carbohydrate molecules ( a six-membered ring for a pyranose and a five-membered ring for a furanose that are composed of carbon atoms and only one oxygen atom ) are identified . For each carbohydrate-like molecule , the chemical groups attached to each position of the ring and their orientations are compared with a pre-defined table to identify the correct chemical name for the carbohydrates . Glycan chains are constructed by examining the glycosidic linkages between the carbohydrate molecules that have chemical bonds between them . As of December 2011 , there were 2 , 517 PDB entries and 10 , 769 N-linked glycan chains in the RCSB database . The glycan fragment structure database , including the substructures of the original N-glycan chains , was generated , which resulted in a total of 48 , 568 N-glycan fragment chains . From the N-glycan fragment database , we have collected glycan structures composed of more than 3 carbohydrate units . A glycan structure was excluded when its resolution was higher than 3 Å or when it had less than 20 structures in total , resulting in the 35 N-glycan fragment sequences listed in Table S1 . An N-glycan structure pair is called “non-homologous” when the sequence similarity of the parent proteins is less than 30% . Because a glycoprotein can have multiple glycosylation sites in a single domain , if the distance between the backbone Cα atoms of the two glycosylated Asn residues is more than 10 Å after alignment of the glycoprotein chains using TM-align [43] , the N-linked glycan structure pairs are considered “non-homologous” glycans . The rest of the N-glycan structure pairs are called “homologous” glycans . Figure 1 summarizes the protocol for building the N-glycan structure dataset . To quantify the conformational variability of the PDB N-glycan structures , it is essential to know the upper bound of the conformational variability in a given oligosaccharide . In protein structural biology , the upper bound of conformational variability is estimated by using the non-homologous protein structure pool and sequence-independent structure alignment methods [48]–[50] . However , because such sequence-independent structure alignment methods are not available for oligosaccharides , it is difficult to estimate the upper bound of the conformational variability in oligosaccharides only using the crystal structures in the PDB . Instead of using the crystal structures directly , a conformational pool that contains diverse conformations of a specific N-glycan sequence was generated as follows . For each of the 35 N-glycan sequences , a total of 1 , 000 , 000 glycan conformations were generated in an iterative fashion . The initial structures were generated by using the IC BUILD command in the CHARMM biomolecular simulation program [51] according to the glycan sequence . For each iteration , a glycosidic linkage was randomly selected and a new torsion angle value was also randomly chosen based on the accessible glycosidic torsion angles of the corresponding glycosidic linkage type . If the newly generated conformation had bad contacts with neighboring atoms , the conformation was rejected and the protocol was repeated until no bad contacts were found . If a conformation had no bad contacts , the conformation was recorded and the protocol repeated until 1 , 000 , 000 conformations were generated . A bad contact was defined by the CHARMM van der Waals energy higher than 10 kcal/mol . Accessible glycosidic torsion angle values were used rather than the values observed in the PDB because the number of observations is limited for certain types of glycosidic linkages . For example , Figure S6 in Supplementary Material shows the resulting glycosidic torsion angle distributions of the N-glycan core sequence using the accessible glycosidic torsion angle values , and Figure S7 shows the torsion angle values observed in the PDB , respectively . To construct an accessible glycosidic torsion angle map , a total of 13 adiabatic ( φ , φ , ω ) potential maps were constructed for each distinct glycosidic linkage type found in the 35 N-glycan sequences . For each glycosidic linkage type , a disaccharide connected by the corresponding glycosidic linkage type was generated by CHARMM [51] , and the CHARMM carbohydrate force field [52]–[54] was used to evaluate the energy . The adiabatic map was generated by evaluating the energy over a grid of glycosidic torsion angles with a grid spacing of 5° , resulting in a total of 373 , 248 grid points for ( 1→6 ) linkages ( φ , φ , ω ) and 5 , 184 grid points for the rest of the glycosidic linkages ( φ , φ ) . At each grid point , the conformations were minimized with the dielectric-screened Coulombic electrostatic and Lennard-Jones potential energy while the glycosidic torsion angles were restrained and a harmonic restraint potential was applied to the carbohydrate rings to prevent the distortion of the ring geometry . The generated adiabatic potential energy map was converted to a torsion angle probability map using the Boltzmann distribution . Finally , the resulting distribution was compared with the glycosidic torsion angles observed in the PDB using the Glycan Fragment DB [55] , available at www . glycanstructure . org . The glycosidic torsion angle probability maps and the observations in the PDB matched well in general . However , the torsion angle probability map was clearly more restricted ( data not shown ) . To remedy the restricted conformational space , glycosidic torsion angle pairs having probability above 0 . 0001 were considered “accessible”; this covers on average about 65% of the observed glycosidic torsion angles in the PDB . The N-glycan structural similarity was measured by calculating pairwise RMSD in the following three different ways: First , the heavy atoms in the carbohydrate ring ( C1 , C2 , C3 , C4 , C5 , and O5 ) were used for the alignment of two N-glycan structures and in the RMSD calculation . Second , to examine the variability of N-glycan orientations with respect to the protein , the heavy atoms of glycosylated Asn residues were used to define the alignment , and then the Euclidean distance of the N-glycan structures was calculated using the carbohydrate ring heavy atoms . Third , many crystal structures only have a few residues at the glycosylation site due to difficulties associated with glycan crystal structure determination , and these partial glycan structures can be used to model the rest of a full glycan structure . To examine the efficacy of such an approach in obtaining a better N-glycan orientation with respect to the protein , the carbohydrate ring heavy atoms of the first two residues were used for the alignment of N-glycan structures , and then the Euclidean distance of the N-glycan structures was calculated using the ring heavy atoms excluding the first two residues . The statistical significance of the structural similarity between two glycan structures was estimated by comparing the structural similarity of 124 , 750 random glycan structure pairs for each N-glycan sequence . The structural similarity of random glycan structure pairs was calculated by the identical procedure described above . Using the statistical model , p-values of the corresponding structural similarity measure can be calculated . This allows us to compare structural similarity across different glycan sequences and lengths . Each RMSD distribution for each glycan sequence was modeled by the generalized extreme value distribution , ( 1 ) where . The variable represents the RMSD of a structure pair; , , and are the location , scale , and shape parameters , respectively . These parameters were obtained through the maximum likelihood estimates by the EVD package in R ( http://www . r-project . org ) . 35 sets of determined parameters are given in Table S2 and the fitting results are shown in Figure S2 . The resulting correlational coefficients ( ) are generally good except for a few sequences . The correlation coefficients improved when more “liberal” protocols were used ( e . g . , when the glycosidic linkage was not restricted and larger energy cutoff values were used to define bad contacts; data not shown ) . However , such protocols may produce unrealistic random glycan conformers and are not used in this work . The p-value of a glycan structure pair from the PDB having RMSD values smaller than the random glycan conformation background was calculated by ( 2 ) The local protein structures are defined for protein residues having any heavy atoms within 6 Å from any glycan heavy atoms . The local protein structures were derived from the PDB structure files in our dataset , and the TM-align algorithm [43] was used to compare the structural similarity of a given local protein structure pair . Any local protein structures having less than 5 residues were excluded . The TM-scores calculated by TM-align were normalized by the length of the smaller structure . To estimate statistical significance , we have derived a random local protein structure pool using the N-linked glycoproteins in the PDB . Briefly , from the PDB , a non-redundant N-linked glycoprotein structure list having at least one carbohydrate residue and protein sequence similarity less than or equal to 30% were generated . A random local protein structure pool was derived from the protein residues having any heavy atoms within 6 Å from any of the carbohydrate heavy atoms . The TM-align algorithm was used to calculate the distribution of TM-scores from the random local protein structure pairs . The calculated TM-score distribution was fit using the generalized extreme distribution ( Eq . ( 1 ) ) , and the p-values of having TM-scores larger than the random background were estimated using Eq . ( 2 ) . Although there are several local structure alignment tools available [56]–[58] , it was difficult to directly utilize them in this study because many of them are highly customized to specific domains , such as a protein-protein interface or protein-ligand interface . Thus , we used TM-align [43] to compare local structure similarity . Although TM-align is not designed to compare local structure similarity , it performed well in our internal testing and correctly found most homologous glycoproteins having similar local protein Cα structures; also see ref for protein local structure comparisons [44] . The residue distance is defined as the minimum number of glycosidic linkages between carbohydrate monomers , including the glycosidic linkage to Asn . For each of 35 N-glycan sequences , three types of internal substructures were generated; a ) residue distance up to 3 , b ) residue distance up to 4 , and c ) residue distance up to 4 , excluding residues linked by the 1–6 linkage . Then , the RMSD of substructures were measured after alignment using the carbohydrate ring atoms in the substructure . To estimate the statistical significance of the internal substructures , the random glycan internal structure pool was generated for each of three different types of substructures . The resulting random background distributions were fit using Eq . ( 1 ) and p-values were calculated using Eq . ( 2 ) .
|
An N-glycan is a carbohydrate chain covalently linked to the side chain of asparagine . Due to the flexibility of carbohydrate chains , it is believed that the N-glycan chains would not have a well-defined structure . However , our survey of N-glycan structures in the PDB shows that the N-glycan structures found on the surfaces of homologous glycoproteins are significantly conserved . This suggests that the interaction between the carbohydrate and the protein structure around the glycan chain plays an important role in determining the N-glycan structure . While the global N-glycan structures found on the surfaces of non-homologous glycoproteins are not conserved , the conformations of the carbohydrate residues that are closer to the protein appear to be more conserved . Our analysis highlights the applicability of template-based approaches used in protein structure prediction to structure prediction and modeling of N-glycans of glycoproteins .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"biology",
"computational",
"biology"
] |
2013
|
Restricted N-glycan Conformational Space in the PDB and Its Implication in Glycan Structure Modeling
|
The innate immune system is the first line of host defense against invading organisms . Thus , pathogens have developed virulence mechanisms to evade the innate immune system . Here , we report a novel means for inhibition of neutrophil recruitment by Group A Streptococcus ( GAS ) . Deletion of the secreted esterase gene ( designated sse ) in M1T1 GAS strains with ( MGAS5005 ) and without ( MGAS2221 ) a null covS mutation enhances neutrophil ingress to infection sites in the skin of mice . In trans expression of SsE in MGAS2221 reduces neutrophil recruitment and enhances skin invasion . The sse deletion mutant of MGAS5005 ( ΔsseMGAS5005 ) is more efficiently cleared from skin than the parent strain . SsE hydrolyzes the sn-2 ester bond of platelet-activating factor ( PAF ) , converting biologically active PAF into inactive lyso-PAF . KM and kcat of SsE for hydrolysis of 2-thio-PAF were similar to those of the human plasma PAF acetylhydrolase . Treatment of PAF with SsE abolishes the capacity of PAF to induce activation and chemotaxis of human neutrophils . More importantly , PAF receptor-deficient mice significantly reduce neutrophil infiltration to the site of ΔsseMGAS5005 infection . These findings identify the first secreted PAF acetylhydrolase of bacterial pathogens and support a novel GAS evasion mechanism that reduces phagocyte recruitment to sites of infection by inactivating PAF , providing a new paradigm for bacterial evasion of neutrophil responses .
Neutrophils are one of the first responders of innate inflammatory cells to migrate towards the site of infecting agents . Evasion of the neutrophil microbicidal response is critical for survival , dissemination , and infectability of bacterial pathogens . Bacterial pathogens evade the neutrophil responses by multiple mechanisms , including inhibition of neutrophil infiltration , antiphagocytosis , and killing of neutrophils . Group A Streptococcus ( GAS ) causes a variety of diseases , ranging from relatively mild pharyngitis to potentially lethal invasive infections , such as necrotizing fasciitis [1] . The success of GAS as a pathogen is based , in part , on its ability to evade the innate immune system . GAS expresses extracellular peptidases ScpA and SpyCEP/ScpC to inhibit neutrophil recruitment by degrading the chemotactic C5a peptide and IL-8/CXC chemokines , respectively [2] , [3] , [4] , [5] . The hyaluronic acid capsule and surface M protein made by GAS confer resistance to opsonophagocytosis and phagocytosis by neutrophils [6] , [7] . Secreted DNase Sda1 helps GAS escape from neutrophil extracellular traps [8] . Mac/IdE inhibits opsonophagocytosis [9] , [10] . Streptolysin S and streptolysin O kill and induce apoptosis of neutrophils [11] , [12] . GAS pathogenesis is mediated by many virulence factors , and alteration in regulation of virulence factors greatly affects clinical outcomes . The two component regulatory system CsrRS/CovRS negatively regulates many virulence factor genes of GAS , including most of the virulence factors involved in the innate immune evasion [13] , [14] . Nonsense and missense mutations in csrRS/covRS occur during human infections and are epidemiologically linked to severe GAS infections [15] . Selection of hypervirulent strains with csrRS/covRS mutations during experimental invasive infections in mice further highlights the critical role of csrRS/covRS mutations in progression of invasive GAS infections [16] , [17] , [18] . Loss of SpeB and enhanced production of the hyaluronic acid capsule contribute to the progression of invasive GAS infections [19] , [20] . Enhanced production of the virulence factors in the innate immune evasion as a result of csrRS/covRS mutations plays a key role in selection for hypervirulent csrRS/covRS mutants . The DNase Sda1 helps GAS escape neutrophil extracellular traps and provides selection pressure for csrRS/covRS mutations [8] . Neutrophil infiltration to infection sites is almost completely inhibited in some necrotizing fasciitis patients and during experimental severe soft tissue infections in primates and mice [2] , [21] , [22] , [23] . Enhanced production of SpyCEP/ScpC and ScpA as a result of csrRS/covRS mutations are believed to contribute to the enhanced inhibition of neutrophil recruitment in severe invasive infections . It is not known whether SpyCEP/ScpC and ScpA are entirely responsible for the dramatic inhibition of neutrophil recruitment by hypervirulent GAS strains with csrRS/covRS mutations . Platelet-activating factor ( PAF ) also has chemotactic activity for inflammatory cells . PAF is a phospholipid mediator with the chemical structure of 1-O-alkyl-2-acetyl-sn-glycero-3-phosphorylcholine [24] . PAF is produced by endothelial cells , neutrophils , macrophages , and eosinophils in responses to proinflammatory cytokines , phagocytosis , and/or other stimuli [25] . This important phospholipid mediator has diverse and potent biological activities , including participation in normal physiological processes , such as inflammation , hemostasis , and reproduction , and contribution to pathological responses , including asthma , ischemia , gastric and pulmonary distress , allergy , and shock [26] . Particularly , PAF can activate platelets [27] and neutrophils [28] . The biological activities of PAF are mediated by a G protein-linked receptor ( PAFR ) that is expressed on the surface of various cell types [29] , [30] . The biological activities of PAF are regulated by PAF acetylhydrolases that hydrolyze the sn-2 acetyl ester bond , converting PAF into acetic acid and lyso-PAF . Four mammal PAF acetylhydrolases , secreted or plasma , two intracellular type I , and intracellular type II PAF acetylhydrolases , have been described [31] , [32] , [33] , [34] . The plasma and intracellular type II PAF acetylhydrolases belong to group VII of phospholipases A2 , and the type I PAF acetylhydrolases are classified as group VIII phospholipases A2 [35] . Group VIII PAF acetylhydrolases are completely specific for PAF whereas the plasma and type II PAF acetylhydrolases hydrolyze unmodified sn-2 fatty acyl residues up to 5 or 6 carbon atoms long and longer sn-2 acyl residues with modification by oxidation [35] . PAF acetylhydrolase activity has been also detected in bacteria and fungus . An intracellular yeast group VII PAF acetylhydrolase enhances the viability of yeast under oxidative stress [36] . The spirochete Leptospira interrogans produces a PAF acetylhydrolase [37] . An apparently intracellular esterase Est13 from an earthworm gut-associated microorganism inhibits PAF-induced platelet aggregation [38] . Both L . interrogans PAF acetylhydrolase and Est13 share sequence homology with the α1 subunit of the type intracellular I mammalian PAF acetylhydrolase . The function of these bacterial PAF acetylhydrolases is not known . These yeast and bacterial PAF acetylhydrolases are intracellular proteins . The esterase secreted by GAS ( designated SsE ) is a protective antigen [39] and is regulated by CsrRS/CovRS and required for GAS virulence and dissemination [40] . The basis for the contribution of SsE to GAS virulence and dissemination is unknown . Identification of the esterase target is essential for elucidating the functional mechanism of SsE . The homologue of SsE in the horse pathogen Streptococcus equi possesses optimal activity to acetyl esters [41] . We hypothesize that SsE targets PAF and is involved in evasion of the innate immune system . Here , we report on studies designed to test this hypothesis . Our findings demonstrate that SsE is indeed a potent PAF acetylhydrolase and is required for inhibition of neutrophil infiltration . We also present evidence for one of mechanisms for SsE to evade the neutrophil response by targeting PAF , identifying a new novel virulence factor for innate immune evasion .
Identification of the esterase target is essential for elucidating the mechanism by which GAS uses SsE to contribute to GAS virulence and dissemination . Since the homologue of SsE in Streptococcus equi has optimal activities to acetyl esters [41] , we considered whether the target of SsE is a molecule with a short-chain acyl ester group . PAF appears to be a good candidate as the target of SsE since it has an acetyl group and is an inflammatory mediator and chemoattractant [28] , [42] . PAF was incubated with SsE , and the reaction was analyzed by thin layer chromatography ( TLC ) , which could resolve PAF and lyso-PAF because PAF migrates much faster ( Figure 1 ) . SsE-treated PAF migrated the same distance as lyso-PAF , indicating that PAF was hydrolyzed by SsE . To confirm that PAF hydrolysis was due to the enzymatic activity of SsE , we performed a control experiment using SsES178A mutant protein . This mutant lacks the catalytic residue , Ser178 , and , therefore , lacks enzymatic activity [39] . Indeed , SsES178A-treated PAF and untreated PAF had the same migration rate . These results indicate that SsE hydrolyzes the acetyl ester bond in PAF , resulting in lyso-PAF . Next , we used liquid chromatography/positive ion electrospray mass spectrometry ( LC-MS ) to confirm SsE-catalyzed hydrolysis of PAF . PAF ( 1 . 4 mM ) was mixed with 80 nM SsE , and an aliquot was taken from the reaction immediately ( 0 min ) or at 5 min after mixing and diluted with an equal volume of acetonitrile to stop the reaction . A control reaction containing PAF and SsES178A was performed under the identical conditions and stopped at 40 min after mixing . The elution times of lyso-PAF and PAF on a C8 column were 4 . 15 and 4 . 38 min , respectively ( Figure 2A ) , and the accurate m/z values of PAF and lyso-PAF were 524 . 3722 and 482 . 3600 , respectively ( Figure 2B ) . We found that 57% and 100% of PAF was converted into lyso-PAF for the SsE-treated PAF samples obtained at 0 and 5 min after mixing , respectively ( Figure 2B and 2C ) , whereas no PAF was hydrolyzed into lyso-PAF at 40 min after mixing PAF with inactive SsES178A ( Figure 2D ) . These results unambiguously demonstrate that SsE catalyzes the conversion of PAF into lyso-PAF . Thus , SsE is a PAF acetylhydrolase . To determine whether SsE can hydrolyze long-chain acyl group at the sn-2 position , we tested whether SsE hydrolyzes heptanoyl thio-PC ( 1-O-hexadecyl-2-heptanoyl glycerol-3-phosphocholine ) , an analogue of 2-thio-PAF , which is used in a colorimetric assay for PAF acetylhydrolases [43] . No hydrolysis of heptanoyl thio-PC was detected , whereas 2-thio-PAF was rapidly hydrolyzed ( Figure 3A ) , indicating that SsE cannot hydrolyze esters with a long-chain acyl group . We also determined whether the PAF acetylhydrolase activity of SsE requires Ca2+ . The observed initial hydrolysis rates were measured in reactions containing 2 . 3 nM SsE and 20 µM 2-thio-PAF in the presence of 0 . 0 mM Ca2+ , 1 . 0 mM EDTA , or 1 . 0 mM Ca2+ . The measured rates of 2-thio-PAF hydrolysis were 7 . 0 , 9 . 2 , and 7 . 8 µM min−1 , respectively . Thus , the activity of SsE does not require Ca2+ and other metal ions that can form a complex with EDTA . These properties of SsE are similar to those of eukaryotic PAF acetylhydrolases . To determine whether SsE is a potent PAF acetylhydrolase , we measured the kcat and KM values of SsE for hydrolysis of 2-thio-PAF using the PAF acetylhydrolase assay kit and compared them with those of recombinant human plasma PAF acetylhydrolase . The initial reaction rates were obtained as described in Figure S1 . The relationship of the observed rates versus 2-thio-PAF concentration fits the Michaelis-Menten equation ( Figure 3B ) , yielding a kcat of 69 . 6 s−1 and an apparent KM of 7 . 0 µM for SsE . In comparison , kcat and KM of recombinant human plasma PAF acetylhydrolase were determined to be 15 . 4 s−1 and 8 . 0 µM , respectively . These measurements indicate that SsE has similar KM with and higher kcat than the human enzyme . PAF has a variety of biological functions , including activation of neutrophils , and the acetyl ester group at sn-2 is critical for its activities . Thus , SsE-catalyzed hydrolysis of PAF should inactivate the functions of PAF . We tested whether treatment of PAF with SsE alters the capacity of PAF to activate neutrophil Ca2+ mobilization . SsE-treated , SsES187A-treated , and untreated PAF and SsE alone were added to human neutrophils preloaded with Fluo-4 acetoxymethyl ester , and changes in fluorescence due to the increase in free intracellular Ca2+ were monitored . SsE-treated PAF at 50 ng/ml and the protein controls were not able to mobilize an intracellular Ca2+ flux , whereas SsES178A-treated and untreated PAF at 0 . 05 ng/ml induced a normal Ca2+ flux ( Figure 4A ) . Thus , SsE abolishes the capacity of PAF to activate this neutrophil response . PAF is also a potent neutrophil chemoattractant . To determine whether SsE could inhibit PAF-induced neutrophil chemotaxis , we assessed the effect of SsE on PAF-induced neutrophil migration . As shown in Figure 4B , PAF was chemotactic for human neutrophils , whereas SsE-treated PAF lost the chemotactic activity , and the number of migrated neutrophils in the presence of PAF that was treated with SsE were similar to those of the buffer and protein only controls . In contrast , treatment with inactive SsES178A did not reduce PAF-induced neutrophil chemotaxis ( Figure 4B ) . These results indicate that SsE inhibits PAF-induced neutrophil chemotaxis and that the inhibition requires SsE enzymatic activity . Since SsE can abolish PAF-induced activation and chemotaxis of neutrophils , we tested whether SsE is involved in innate immune evasion during GAS infections . We first examined the infection sites and performed histological analyses . MGAS5005 extensively spreads from inoculation sites by 24 hours after inoculation ( Figure S2A ) whereas the ΔsseMGAS5005 mutant remained at the inoculation site ( Figure S2B ) . The histological analyses of the skin infection sites with the Gram and hematoxylin and eosin ( H&E ) stains reveal distinct patterns of inflammatory cell infiltration between MGAS5005 and ΔsseMGAS5005 sites at 24 h after inoculation . Inflammatory cells and amorphous materials were kept away from GAS at the MGAS5005 inoculation site , and few neutrophils were found at the spread area of MGAS5005 ( Figure S3A and S3B ) . In contrast , inflammatory cells were present throughout the inoculation site with more inflammatory cells surrounding the infection site ( Figure S3C and S3D ) . The distinct details of these patterns are more evident at a higher magnification . There are five morphological zones at an end of the MGAS5005 inoculation site starting from the interior side of the skin ( the right side in panels A and B of Figure 5 ) : Zone 1 , neutrophils and other inflammatory cells without GAS; Zone 2 , amorphous host materials lack of GAS; Zone 3 , a few inflammatory cells that could reach the boundary of the GAS territory were victimized by and associated with massive amount of GAS; Zone 4 , necrotized adipose tissue and GAS without inflammatory cells; and Zone 5 , invasion of GAS along the interstitial space of the adipose cells ( Figure 5A and 5B ) . Thus , MGAS5005 not only reduces infiltration of neutrophils but also keep inflammatory cells away . However , inflammatory cells and ΔsseMGAS5005 bacteria were mingled throughout the infection site ( Figure 5C and 5D ) . Similar results were obtained in CD-1 Swiss mice . Next , we used the myeloperoxidase assay [44] to quantify neutrophil ingress to the skin infection sites of MGAS5005 and ΔsseMGAS5005 at 24 h after subcutaneous infection of BALB/c mice . The mean neutrophil number ± SD of the Δsse infection site was ( 1 . 1±0 . 12 ) ×106/mm2 , which was 19 . 6 and 346-fold greater than the neutrophil number at the MGAS5005 inoculation site [ ( 5 . 4±2 . 3 ) ×104 neutrophils/mm2] and at the spread infection area of MGAS5005 [ ( 3 . 1±0 . 87 ) ×103 neutrophils/mm2] . Reverse complementation of ΔsseMGAS5005 with the sse gene ( Δsse-sse ) restored the inhibition of neutrophil recruitment [ ( 5 . 6±1 . 0 ) ×104 neutrophils/mm2] . The difference is significant between the ΔsseMGAS5005 sample and each of the other samples but insignificant among the other samples in one way ANOVA analysis using the Tukey's Multiple Comparison Test ( Figure 6A ) . The receptor of PAF ( PAFR ) is a G protein-coupled receptor that mediates the biological activities of PAF . We used PAFR-deficient mice [30] to test whether SsE inhibits neutrophil infiltration by hydrolyzing PAF . MGAS5005 induced low and similar levels of neutrophil recruitment in both BALB/c and PAFR−/− mice . However , the mean number of recruited neutrophils at the ΔsseMGAS5005 infection site was reduced by 47% in PAFR−/− mice compared with BALB/c mice ( Figure 6B ) . The reduction of neutrophil influx due to the absence of the PAF receptor was 52 . 7% of the enhancement of neutrophil influx as a result of the sse deletion . These results suggest that targeting PAF by SsE is an equally important mechanism as an PAF-independent mechanism . These results strongly suggest that PAF plays a significant role in neutrophil infiltration in GAS infections and that SsE-mediated hydrolysis of PAF contributes to the observed reduction in neutrophil infiltration . Since Δsse bacteria were associated with high levels of neutrophils , these bacteria should be killed by recruited neutrophils . Indeed , the numbers of viable ΔsseMGAS5005 at 24 and 48 hours post-inoculation were 8 . 3% and 4 . 8% of those found at 1 h after inoculation , respectively; whereas the numbers of MGAS5005 at 24 and 48 h post-inoculation were 70% and 128% of those found at 1 h after inoculation , respectively ( Figure 6C ) , suggesting that ΔsseMGAS5005 is cleared more efficiently than MGAS5005 at skin infection sites . In a transcription profiling analysis for MGAS5005 and ΔsseMGAS5005 using the NimbleExpress Streptococcus pyogenes arrays , the transcription levels of the genes regulated by the multiple gene regulator of GAS ( Mga ) and CsrRS/CovRS in ΔsseMGAS5005 were 70% to 135% of those in MGAS5005 at the mid-exponential growth phase except that sse transcript was not detected in ΔsseMGAS5005 ( Figure S4 ) . These results rule out the possibility that the phenotype of ΔsseMGAS5005 is caused by alteration in transcription of the scpA , spyCEP/scpC , sda1/sdaD2 , slo , sagA , hasA , speB , and emm genes , which are involved in innate immune evasion by GAS . MGAS5005 has a natural null covS deletion , which enhances expression of sse and many other virulence genes [18] , [40] . To test whether SsE contributes to pathogenesis and inhibition of neutrophil recruitment in GAS with the wild-type csrRS/covRS genes , we deleted the sse gene in MGAS2221 . Fifty seven percent of BALB/c mice infected subcutaneously with 1 . 5×108 cfu of MGAS2221 were dead whereas all mice infected with 1 . 6×108 cfu ΔsseMGAS2221 survived ( P = 0 . 0218 ) ( Figure 7A ) . In a separate experiment , 3 . 9×107 cfu MGAS2221 or 1 . 6×108 cfu ΔsseMGAS2221 bacteria were inoculated into BALB/c mice . The lesion appearance was obviously different between the wt and mutant infection sites ( Figure 7B ) . The number of neutrophils at the ΔsseMGAS2221 site was significantly higher than that at the MGAS2221 site ( mean neutrophil number ± SD: ΔsseMGAS2221 , ( 2 . 4±1 . 4 ) ×105/mm2; MGAS2221 , ( 1 . 2±0 . 6 ) ×105/mm2 ) ( P = 0 . 0420 ) ( Figure 7C ) . Conversely , the size of the ΔsseMGAS2221 site was significant smaller than that of the MGAS2221 site ( mean size ± SD: MGAS2221 , 106±20 mm2; ΔsseMGAS2221 , 77±6 mm2 ) ( P = 0 . 0014 ) ( Figure 7D ) . It should be stressed that the significant role of SsE in the invasion of skin tissue and inhibition of neutrophil recruitment was observed with a dose of ΔsseMGAS2221 that was 3 times higher than that of MGAS2221 . The results using the higher dose of the mutant suggest that the mutant phenotype is not caused by a growth defect . Thus , SsE can reduce neutrophil recruitment and enhances soft tissue invasion in infection with a representative M1T1 strain with the wild-type csrRS/covRS background . The effects of in trans expression of SsE on neutrophil recruitment and lesion size during subcutaneous MGAS2221 infection of mice further confirm the role of SsE in inhibition of neutrophil recruitment and enhancement of soft tissue invasion by GAS . The sse gene was cloned into pDCBB [45] , yielding pSsE . At the early growth phase ( OD600 = 0 . 2 ) , SsE was detected in the supernatant of MGAS2221/pSsE but not MGAS2221/pDCBB ( vector control ) by Western blotting analysis , whereas the secreted protein Spy0019 was detected at similar levels in the supernatant of both strains ( Figure 7E ) . These results indicate that the introduction of pSsE into MGAS2221 enhances SsE production . In trans production of SsE increased lesion size by 124% compared with the vector control ( Lesion size ± SD: MGAS2221/pDCBB , 41±15 mm2; MGAS2221/pSsE , 92±25 mm2 ) ( P = 0 . 0047 ) ( Figure 7F ) . Inversely , in trans production of SsE reduced neutrophil recruitment by 72% ( mean neutrophil number ± SD: MGAS2221/pDCBB , ( 6 . 2±0 . 28 ) ×105/mm2; MGAS2221/pSsE , 1 . 7±0 . 11 ) ×105/mm2 ) ( P = 0 . 0111 ) ( Figure 7G ) . The ΔsseMGAS5005 mutant has a longer early growth phase by about 15 min ( Figure 8A ) and about 10% more viable CFU per OD600 at the exponential growth phase ( data not shown ) than its parent strain in Todd-Hewitt broth supplemented with 0 . 2% yeast extract ( THY ) . Consistent with this result , in trans overexpression of SsE in MGAS2221 shows a 20-min shorter early growth phase than the vector control ( Figure 8B ) . However , MGAS2221 and ΔsseMGAS2221 have identical growth curves in THY ( Figure 8C ) . Thus , the effect of sse expression on the length of early growth phase is obvious when SsE is highly produced . To examine the growth of the mutants in vivo , we performed a competitive growth assay using an air sac infection model . A 1∶1 ΔsseMGAS2221∶MGAS2221 or ΔsseMGAS5005∶MGAS5005 mixture was injected with air in the subcutis of mice , and , 24 h later , the air sac was lavaged after the mice were euthanized . The lavage samples were plated , and the Δsse∶wt GAS ratio of the GAS colonies was determined by PCR analysis . The Δsse∶wt GAS ratio in the inoculum was measured by plating the individual GAS suspension prior to mixing . The mean number of MGAS5005 and MGAS2221 at 24 h was 11 and 3 times as the corresponding number at 8 h , respectively ( Figure 8D ) , indicating that GAS grew in the air sac . The competitive index , the Δsse∶wt ratio in the lavage sample/the ratio in the inoculum , for both ΔsseMGAS2221 and ΔsseMGAS5005 has a mean value of about 1 ( Figure 8E ) , indicating that each mutant and its parent strain have similar growth in vivo . These data indicate that the phenotype of ΔsseMGAS5005 and ΔsseMGAS2221 is not caused by a growth phenotype .
This study presents three major findings regarding evasion of the innate immune system by GAS . First , SsE significantly contributes to GAS inhibition of neutrophil recruitment . Second , SsE is a potent PAF acetylhydrolase and the first secreted bacterial PAF acetylhydrolase identified so far . Third , SsE inactivates the ability of PAF to induce activation and migration of neutrophils , and the PAF receptor significantly contributes to neutrophil recruitment in skin GAS infection . These findings identify a new means for evasion of the innate immune system by GAS and support a novel paradigm for bacterial inhibition of neutrophil recruitment and function in which neutrophil recruitment is reduced by inactivating PAF . One conclusion of this work is that SsE is required for the severe inhibition of neutrophil recruitment by MGAS5005 in the mouse model of necrotizing fasciitis . This nearly complete inhibition of neutrophil infiltration is similar to that of severe GAS infections in some human patients and experimental animal infections [2] , [21] , [22] , [23] . In addition , SsE is critical for the virulence and dissemination of MGAS5005 and is a protective antigen [39] , [40] . SsE also reduces neutrophil recruitment and enhances virulence and skin tissue invasion in infection with MGAS2221 . Thus , SsE is a significant contributor to the innate immune evasion and tissue invasion by GAS with or without covRS mutations . It is well known that GAS produces C5a peptidase ScpA and IL-8/CXC peptidase SpyCEP/ScpC to reduce neutrophil recruitment . SpyCEP/ScpC reduces neutrophil infiltration in soft tissue infections of mice [2] , [3] , promotes resistance to neutrophil killing [5] and GAS dissemination [46] , [47] , and alters pathogenesis [48] . Immunization with ScpA prevents nasopharyngeal GAS colonization of mice [49] . GAS also produces virulence factors , such as the hyaluronic acid capsule , M protein , streptolysins S and O , opsonophagocytosis inhibitor Mac/IdeS , and DNases , to cripple the innate immune system . Our work adds a new virulence factor to the large array of GAS virulence factors that interfere with the bactericidal function of neutrophils . Another conclusion of this work is that SsE contributes to the enhanced inhibition of neutrophil recruitment as a result of the null covS mutation in MGAS5005 . MGAS5005 has a genetic makeup almost identical with that of MGAS2221 , displaying 7 synonymous and 9 non-synonymous single nucleotide alterations , two single base mutation , and presence of an IS element [18] . However , MGAS5005 has a null covS deletion but MGAS2221 has the wild-type csrRS/covRS genes . Alteration of the transcription of the CsrRS/CovRS-regulated genes by the null covS mutation is apparently the cause for the lower neutrophil recruitment in MGAS5005 infection than in MGAS2221 infection . Expression of the sse gene is enhanced by 30 fold by the covS null mutation in MGAS5005 [40] . Deletion of the sse gene in MGAS5005 does not dramatically change expression of the CsrRS/CovRS- and Mga-regulated virulence genes but reversed the covS mutation-induced reduction of neutrophil infiltration . Thus , the enhanced production of SsE is an critical factor for the increase in inhibition of neutrophil recruitment and soft tissue infection during MGAS5005 infection in comparison with MGAS2221 infection . This conclusion is supported by the decrease in neutrophil recruitment and increase in skin invasion that are caused by in trans production of SsE in MGAS2221 . SpyCEP/ScpC and ScpA are also up-regulated as a result of covS null mutations [45] . We propose that SsE , SpyCEP/ScpC , and ScpA can all reduce neutrophil recruitment during infections of GAS with the wild-type covRS genes and cripple neutrophil infiltration when they are highly produced as a result of null covS mutations . The requirement of SsE in the inhibition of neutrophil infiltration and invasion of skin tissue by MGAS5005 indicates that enhanced production of SsE , like enhancement in capsule production and suppression of SpeB production , is critical for covS mutations-mediated progression of invasive GAS infection . SsE is a novel , potent bacterial PAF acetylhydrolase . Hydrolysis of PAF by SsE was clearly demonstrated by TLC . Analysis of the SsE/PAF reaction by LC-MS not only confirmed PAF hydrolysis but also demonstrated that SsE-catalyzed PAF hydrolysis was rapid . The PAF acetylhydrolase activity of SsE appears at least to be as potent as the human plasma PAF acetylhydrolase . PAF acetylhydrolase activity has also been detected in bacteria and yeast [36] , [37] , [38] . While the yeast PAF acetylhydrolase enhances the viability of yeast under oxidative stress , the function of L . interrogans PAF acetylhydrolase and Est13 is not known . There is a difference between SsE and the yeast , L . interrogans , and Est13 PAF acetylhydrolases in cellular location . SsE is a secreted protein [39] but the fungus and other bacterial PAF acetylhydrolases described so far are intracellular proteins [36]–[38] . This difference in the cellular location dictates whether these PAF acetylhydrolases can target host PAF . SsE would be able to degrade host PAF produced in response to infection but the fungus and other bacterial PAF acetylhydrolases should not be able to target host PAF . Thus , we have identified the first secreted PAF acetylhydrolase that can target host PAF . SsE has homologues in both Gram-positive and Gram-negative pathogens , such as Streptococcus agalactiae , Streptococcus equi , Streptococcus zooepidemicus , Staphylococcus aureus , Streptobacillus moniliformis , and Actinomyces coleocanis [41 , BLAST results not shown] . The function and functional mechanism of SsE may be relevant to other bacterial infections . Neutrophil infiltration is significantly reduced in ΔsseMGAS5005 but not in MGAS5005 infection in the PAFR−/− mice compared with those in the control mice . These results support a novel mechanism of innate immune evasion: SsE hydrolyzes PAF to reduce neutrophil recruitment . The reduction of neutrophil recruitment to ΔsseMGAS5005 in PAFR−/− is 52% of the enhancement of neutrophil recruitment caused by the sse deletion in MGAS5005 . Thus , the PAF-dependent mechanism is a significant but not only mechanism for SsE to contribute to inhibition of neutrophil recruitment . The role of SsE in skin invasion and inhibition of neutrophil recruitment appears not to be caused by a growth phenotype . First , ΔsseMGAS2221 and MGAS2221 have similar growth both in vitro and in vivo , and , thus , the phenotype of ΔsseMGAS2221 in neutrophil recruitment , skin invasion and virulence is not caused by a growth phenotype . Furthermore , ΔsseMGAS2221 at a dose 4 times higher than that of MGAS2221 displayed the Δsse phenotype . Third , although ΔsseMGAS5005 has a longer early growth phase than the parent strain , the two strains have similar growth in vivo . Fourth , the decrease in neutrophil recruitment during the infection of PAFR−/− mice with ΔsseMGAS5005 cannot be explained by a growth phenotype . Finally , immunization of mice with SsE reduces skin invasion by MGAS5005 [39] . Furthermore , the Δsse phenotype is apparently caused by the loss of SsE but not through an indirect effect since the sse deletion did not alter the expression of CsrRS/CovRS- and Mga-regulated virulence factors . ΔsseMGAS5005 has lower cfu numbers than that of MGAS5005 after 4-h incubation in serum [40] . However , this difference in growth in serum between MGAS5005 and ΔsseMGAS5005 is not reflected in the air sac competitive growth assay . The different results in the two assays could depend on the effect of SsE on the early growth phase . High levels of SsE production as a result of covS mutation or in trans overexpression apparently shorten the early growth phase but did not change the doubling time in vitro . The effect of SsE on the length of the early growth phase might be the reason for the difference in cfu of ΔsseMGAS5005 and MGAS5005 in serum because a low dose of bacteria ( 105 cfu ) were inoculated in the serum growth assay . The early growth phase of ΔsseMGAS5005 in the air sac assay may be shortened because nearly 1 , 000-fold more ΔsseMGAS5005 was inoculated in the air sac assay . At the same time , the early growth phase of MGAS5005 in the air sac assay could be longer than that in the serum growth assay because the nutrient in the air sac assay should be less abundant than in serum . Besides the effect on the early growth phase , the yield of ΔsseMGAS5005 in chemically defined medium is lower than that of MGAS5005 [40] , suggesting that SsE may be able to recycle metabolites or surface structures . These differential in vitro growth features of ΔsseMGAS5005 and MGAS5005 appear not to be displayed in vivo , suggesting that the in vitro difference does not represent a genuine growth defect . Nonetheless , the in vitro growth data indicate that SsE can act on the GAS bacteria . This action could be the basis for a PAF-independent mechanism , in addition to the PAF-dependent mechanism , for the innate immune evasion by SsE . Neutrophil influx to ΔsseMGAS5005 sites in the PAFR−/− mice was half of that in the control mice . This is the first demonstration for the importance of the PAF receptor in neutrophil recruitment in response to a bacterial infection . The PAF receptor is not critical for neutrophil infiltration in pulmonary Klebsiella pneumonia , Pseudomonas aeruginosa , Streptococcus pneumoniae infections and polymicrobial sepsis caused by cecum ligation and puncture [50] , [51] , [52] , [53] . This difference suggests that PAF may play a critical role in neutrophil recruitment in skin infection but not in pulmonary infections . It is also possible that these pathogens , like MGAS5005 , can inactivate PAF . Hermoso et al . have found that the protein Pce of Streptococcus pneumoniae hydrolyzes the phosphocholine group of PAF and hypothesized that Pce has the capacity to interact with and hydrolyze PAF in the bloodstream in vivo , impacting on pathogenesis [54] . Apparently , bacterial pathogens have evolved different enzymatic activities to eliminate PAF , supporting an important role of PAF in host responses against bacterial infections . PAF can be involved in innate immune responses in different ways . Administration of PAF can lead to neutrophil infiltration in the lung and skin [55] , and PAF may participate in the inflammatory responses during GAS infections . IL-12-induced chemotaxis of NK cells and neutrophils is mediated by PAF [56] . PAF can activate neutrophils and induce migration of isolated neutrophils [28] . Treatment of PAF with SsE abolishes the ability of PAF to activate and induce migration of neutrophils . PAF can function as a chemoattractant in the neutrophil responses during GAS infection . It is also possible that PAF plays a role in both the inflammatory response and chemotaxis during GAS infection . PAF also activates platelets in human and some animals . However , the inhibition of the PAF-induced activation of platelets does not play a role in the phenotype of the Δsse mutants in the mouse infections since murine platelets do not produce the PAF receptor according to Dr . Guy Zimmerman at University of Utah . We will examine how PAF contributes to the neutrophil response during GAS infections in our follow-up studies .
All animal experimental procedures were 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 protocols for the experiments performed at Montana State University ( MSU ) and Federal University of Minas Gerais ( UFMG ) were approved by the Institutional Animal Care and Use Committee at MSU ( Permit number: 2009-09 ) and the Animal Ethics Committee of Instituto de Ciências Biológicas ( Permit number: 168/11 ) ( Belo Horizonte , Brazil ) , respectively . Blood was collected from healthy donors in accordance with a protocol approved by the Institutional Review Board at MSU ( Protocol No . BL031109 ) . Written informed consent was provided by study participants and/or their legal guardians . PAF ( 1-O-hexadecyl-2-acetyl-sn-glycero-3-phosphorylcholine ) , lyso-PAF C-16 ( 1-O-hexadecyl-sn-glycero-3-phosphocholine ) , human recombinant plasma PAF acetylhydrolase , heptanoyl thio-PC , and the PAF acetylhydrolase assay kit using 2-thio-PAF as the substrate were purchased from Cayman Chemical ( An Harbor , MI , USA ) . Whatman LK6D Silica Gel 60A thin layer chromatography plates were purchased from Whatman International LLC ( Clifton , NJ , USA ) . Recombinant wild-type and S178A mutant SsE proteins were prepared , as previously described [39] . MGAS5005 is a hypervirulent M1T1 GAS strain isolated from an invasive case in Ontario [9] . MGAS2221 is a M1T1 GAS strain isolated from a scarlet fever patient [48] . MGAS5005 and MGAS2221 have almost identical genetic contents but the former has a null covS 1-bp deletion [18] . ΔsseMGAS5005 , an in-frame sse deletion mutant of MGAS5005 missing amino acids 55–261 of SsE and Δsse-sse , a reverse complement strain of Δsse , have been described [40] . The same sse deletion procedure was followed to obtain ΔsseMGAS2221 . These bacteria for experiments conducted at MSU were grown to mid-exponential phase at 37°C in 5% CO2 in THY . GAS bacteria used in the PAFR−/− experiment at UFMG were grown in brain heart infusion broth ( BHI ) . Tryptose agar with 5% sheep blood , THY agar , and BHI agar were used as the solid media . GAS bacteria used for the animal experiments were harvested at the exponential growth phase and washed three times with and resuspended in pyrogen-free phosphate-buffered saline ( PBS ) to desired doses . SsE-catalyzed hydrolysis of PAF was monitored by TLC and LC-MS analyses and a colorimetric assay . For TLC analysis , 1 . 4 mM PAF was mixed with 0 . 08 µM wild-type SsE or SsES178A in 50 µl of 20 mM Tris-HCl , pH 8 . 0 , and the reaction was stopped by adding 50 µl acetonitrile containing 1% formic acid after 10-min incubation at room temperature . Two µl of the reaction samples , untreated PAF , lyso-PAF , and PAF/lyso-PAF mixtures were spotted on a TLC plate , and these compounds were resolved using a methanol/chloroform/water ( 65∶30∶6 by volume ) mixture as the mobile phase . After chromatography , PAF and lyso-PAF were visualized by spraying with 5% ammonium molybdate sulfate and heating . Protein concentrations were determined using the modified Lowry protein assay kit from Pierce with bovine serum albumin as a standard . For LC-MS analysis , PAF hydrolysis reactions were performed as in the TLC analysis and stopped at 0 and 5 min after mixing for the wild-type SsE/PAF reaction and 40 min for the SsES178A/PAF reaction . The samples were diluted with 5% acetonitrile containing 1% formic acid , and 1 µl of the diluted samples were analyzed by reverse-phase liquid chromatography and positive ion mass spectroscopy using an Agilent 1100 HPLC with autosampler ( Agilent Technologies , Inc . , Santa Clara , CA , USA ) and a Bruker micrOTOF mass spectrometer ( Bruker Daltonik GmbH , Bremen , German ) . The reverse-phase chromatography consisted of a 3 . 2-ml gradient between H2O and 95% acetonitrile , both with 0 . 1% formic acid , using a Michrom Bioresources C8 column ( 8×1 mm ) . The LC/MS data were analyzed using DataAnalysis 4 . 0 software ( Bruker Daltonik GmbH ) . The mass spectrometer was calibrated using the peaks between 118 and 922 m/z of the Agilent G2421A electrospray calibrant solution infused directly to the source at a rate of 180 µl/h . PAF and lyso-PAF compounds were identified via high mass accuracy with positive control samples with m/z values of 482 . 3600 and 524 . 3722 , respectively ( actual masses of 482 . 3605 and 524 . 3711 , errors of −1 ppm and +0 . 2 ppm , respectively ) . PAF and lyso-PAF were evaluated for carry-over on the C8 column with blank runs , but the C8 column with the described chromatography had no detectable carry-over between runs . The colorimetric assay used the PAF acetylhydrolase assay kit from Cayman Chemical . The reactions were initiated by mixing 100 µl 20 mM Tris-HCl , pH 8 . 0 , containing 2-thio-PAF at various concentrations and 100 µl Tris-HCl containing 4 . 3 nM SsE and 0 . 5 mM DTNB at room temperature in a 96-well plate . Absorbance at 414 nm ( A414 ) of the reaction mixture was recorded every 6 s using a SPECTRAMax 384 Plus spectrophotometer ( Molecular Devices , Sunnyvale , CA , USA ) and was used to determine initial rates of hydrolysis of 2-thio-PAF as described in the Results section . Neutrophils were isolated from the blood using dextran sedimentation , followed by Histopaque 1077 gradient separation and hypotonic lysis of red blood cells , as described previously [57] . Isolated neutrophils were washed twice and resuspended in HBSS without Ca2+ and Mg2+ for Ca2+ mobilization or with Ca2+ and Mg2+ for chemotaxis measurement . Neutrophil preparations were >95% pure , as determined by light microscopy , and >98% viable , as determined by trypan blue exclusion . Changes in free intracellular [Ca2+] were measured with a FlexStation II Scanning Fluorometer ( Molecular Devices ) using Fluo-4 acetoxymethyl ester ( Invitrogen ) , as previously described [58] . Briefly , human neutrophils , suspended in Hanks' balanced salt solution ( HBSS ) containing 10 mM HEPES , were loaded with Fluo-4 acetoxymethyl ester dye ( 1 . 25 ìg/ml final concentration ) for 30 min in the dark at 37°C . After dye loading , the cells were washed with HBSS containing 10 mM HEPES , resuspended in HBSS containing 10 mM HEPES and Ca2+ and Mg2+ , and separated into aliquots , which were inserted into the wells of flat-bottomed , half-area-well black microtiter plates ( 2×105 cells/well ) . After addition of untreated or SsE-treated PAF , changes in fluorescence were monitored ( λex = 485 nm , λem = 538 nm ) every 5 s for 240 s at room temperature . The chemotaxis assay was performed using the ChemoTx Disposable Chemotaxis System in a 96 well microplate format ( Neuro Probe , Inc . , Gaithersburg , MD , USA ) and the CellTitr-Glo Luminescent Cell Viability Assay ( Promega , Madison , WI , USA ) , as described previously [58] . PAF ( 1 . 4 mM ) was incubated with 0 . 08 µM SsE or SsES178A in 50 µl PBS , pH 7 . 0 , at room temperature for 30 min , and the reaction was stopped by adding an equal volume of acetonitrile . Untreated and treated PAF were diluted to desired concentrations with HBSS containing 10 mM HEPES , Ca2+ , Mg2+ , and 0 . 1% BSA ( HBSS/BSA ) . The protein control reaction samples were diluted by the same fold of the dilution as the treated PAF samples . The samples were added to wells of the assay plate at 30 µl/well in 4 replicates . The plate was covered with the filter , and 4×104 neutrophils/well were placed on the top of the filter . The plate was incubated at 37°C for 1 h . Neutrophils that did not migrate were removed , and 20 µl/well of 2 . 5 mM EDTA was added . After incubating for 10 min at 4°C , the EDTA solution was removed , the plate was centrifuged at 600 rpm for 5 min , and 20 µl/well of CellTitr-Glo Luminescent Cell Viability Assay reagent was added . Luminescence from each well was monitored using a Fluoroscan Ascent FL Luminometer ( Thermo Electron Corporation ) . The number of migrated cells was determined based on a standard curve using known numbers of neutrophils . Groups of five-week-old female inbred BALB/c and outbred CD-1 Swiss mice ( Charles River Laboratory ) were subcutaneously infected with 0 . 2 ml of an OD600 of 0 . 8 of GAS suspension in PBS or at indicated doses . Inocula were determined by plating . Mice were sacrificed at 24 h to collect skin samples for histological analyses and measurement of neutrophil infiltration and GAS CFU . Infected mice in virulence studies were monitored twice a day to get survival rates . The PAFR−/− mouse experiment was similarly performed at Dr . Mauro Teixeira's laboratory at UFMG , Brazil . BALB/c mice ( 8 to 12 week-old ) were obtained from CEBIO ( Bioterism Center ) of UFMG , and PAFR−/− mice ( 8 to 12 week-old ) were generated as previously described and backcrossed at least 10 generations into a BALB/c background [30] , [53] . Mice were housed in standard conditions and had free access to commercial chow and water . Whole infection area in the skin was recognized by the boundary of the inflammation after the skin around the infection site was peeled off ( Figure S2 ) . The skin containing the infection area was excised and traced on a paper , which was used to measure the area of infection sites by weighing the traced paper . Numbers of recruited neutrophils in the excised skin were estimated by the myeloperoxidase assay , as described previously [44] . Skin samples were grinded in 0 . 5% hexadecyltrimethylammonium bromide in 50 mM potassium and sonicated on ice for 15 seconds to extract myeloperoxidase . The samples were frozen and thawed for 3 times , sonicated , and centrifuged at 16 , 000 g for 5 min . The myeloperoxidase activity in the supernatant obtained was measured colormetrically in 0 . 2 ml of 50 mM phosphate buffer , pH 6 . 0 , containing the supernatant , 0 . 167 mg/ml o-dianisidine dihydrochloride , and 0 . 001% hydrogen peroxide . The change in absorbance at 460 nm ( ΔA460 ) was recorded with time with a SPECTRAmax 384 Plus spectrophotometer ( Molecular Devices ) . The myeloperoxidase activity , ΔA460/min , was converted into the number of neutrophils using a stand curve of myeloperoxidase activities versus known numbers of murine neutrophils , which were isolated from the bone marrow of mice , as previously described [57] . Skin samples were excised with a wide margin around the infection site after the skin was peeled off and fixed in 10% neutral buffered formalin for 24 h . The samples were dehydrated with ethanol , cleared with xylene , and infiltrated with paraffin using a Tissue Embedding Console System ( Sakura Finetek , Inc . ) . The paraffin blocks was processed to obtain 4-µm sections , which were stained with H&E or with a tissue Gram stain kit from Richard-Allan Scientific according to the manufacturer's protocol . The stained slides were examined using a Nikon ECLIPSE 80i microscope . Clearance of GAS in the skin was measured by determining the numbers of viable GAS at infection sites . The skin around the infection sites was peeled off , excised , and grinded in 2 ml of PBS to recover bacteria , and the samples at appropriate dilution were plated on tryptose agar with 5% sheep blood to count cfu as the number of viable GAS . In the competitive growth assay , 0 . 2 ml of a 1∶1 ΔsseMGAS5005∶MGAS5005 or ΔsseMGAS2221∶MGAS2221 mixture with 0 . 8 ml air was injected subcutaneously into mice . The mice were euthanized at 24 h after inoculation , and the air sac was lavaged with 1 ml PBS . The lavage samples at appropriate dilution were plated on THY agar plates . The Δsse/wt GAS ratio in the lavage samples was determined by analyzing 96 colonies of each sample with colony PCR using primers 5′-ATAACATTTACATTAAGGAGATAC-3′ and 5′-CAGATTTGGTGTTTGAAAAAG-3′ , which yielded 1232-bp and 611-bp PCR products for the wt and Δsse strains respectively . The Δsse/wt GAS ratio in the inoculum was determined by plating the individual GAS suspension prior to mixing . The competitive index is calculated by dividing the Δsse/wt GAS ratio in the lavage samples by the ratio in the inoculum . The sse gene of MGAS5005 was PCR cloned into pDCBB [45] at the XbaI and EcoRI sites using primers 5′-ATCTAGAATAACATTTACATTAAGGAGATAC-3′ and 5′-AGAATTCCAGATTTGGTGTTT-3′ , yielding pSsE . MGAS2221 was transformed with pSsE for in trans SsE overexpression and with pDCBB for vector control . Levels of SsE in the culture supernatant of MGAS2221/pDCBB and MGAS2221/pSsE were compared using Western blotting , as previously described [39] . Statistic analyses of the data of the animal experiments were performed using the GraphPad Prism software with the following tests: Log-rank ( Mantel-Cox ) Test for the survival data in Figure 7A; one way ANOVA Tukey's Multiple Comparison Test for the data in Figure 6; and one-tailed , unpaired t test for Figures 7C , 7D , 7F , 7G , and 8D .
|
GAS is a major human pathogen causing a variety of infections , including pharyngitis and necrotizing fasciitis . GAS pathogenesis is mediated by a large array of secreted and cell-surface virulence factors . However , the functions of many GAS virulence factors are poorly understood . Recently , we reported that the esterase secreted by GAS ( SsE ) is a CovRS ( the control of virulence two component regulatory system ) -regulated protective antigen and is critical for spreading in the skin and systemic dissemination of GAS in a mouse model of necrotizing fasciitis . This report presents three major findings regarding the function and functional mechanism of SsE: 1 ) SsE contributes to GAS inhibition of neutrophil recruitment; 2 ) SsE is a potent PAF acetylhydrolase and the first secreted bacterial PAF acetylhydrolase identified so far; and 3 ) the PAF receptor significantly contributes to neutrophil recruitment in skin GAS infection . These findings support a novel mechanism for evasion of the innate immune system by GAS that may be relevant to other infections .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"biochemistry",
"infectious",
"diseases",
"bacterial",
"diseases",
"immunity",
"enzymes",
"biology",
"microbiology",
"bacterial",
"pathogens"
] |
2012
|
Group A Streptococcus Secreted Esterase Hydrolyzes Platelet-Activating Factor to Impede Neutrophil Recruitment and Facilitate Innate Immune Evasion
|
Paragonimiasis is a food-borne trematode infection acquired by eating raw or undercooked crustaceans . It is a major public health problem in the far East , but it also occurs in South Asia , Africa , and in the Americas . Paragonimus worms cause chronic lung disease with cough , fever and hemoptysis that can be confused with tuberculosis or other non-parasitic diseases . Treatment is straightforward , but diagnosis is often delayed due to a lack of reliable parasitological or serodiagnostic tests . Hence , the purpose of this study was to use a systems biology approach to identify key parasite proteins that may be useful for development of improved diagnostic tests . The transcriptome of adult Paragonimus kellicotti was sequenced with Illumina technology . Raw reads were pre-processed and assembled into 78 , 674 unique transcripts derived from 54 , 622 genetic loci , and 77 , 123 unique protein translations were predicted . A total of 2 , 555 predicted proteins ( from 1 , 863 genetic loci ) were verified by mass spectrometric analysis of total worm homogenate , including 63 proteins lacking homology to previously characterized sequences . Parasite proteins encoded by 321 transcripts ( 227 genetic loci ) were reactive with antibodies from infected patients , as demonstrated by immunoaffinity purification and high-resolution liquid chromatography-mass spectrometry . Serodiagnostic candidates were prioritized based on several criteria , especially low conservation with proteins in other trematodes . Cysteine proteases , MFP6 proteins and myoglobins were abundant among the immunoreactive proteins , and these warrant further study as diagnostic candidates . The transcriptome , proteome and immunolome of adult P . kellicotti represent a major advance in the study of Paragonimus species . These data provide a powerful foundation for translational research to develop improved diagnostic tests . Similar integrated approaches may be useful for identifying novel targets for drugs and vaccines in the future .
Paragonimiasis is an important food-borne trematode infection ( and a “neglected tropical disease” ) that is caused by lung flukes in the genus Paragonimus [1]–[3] . More than 50 Paragonimus species have been described , and nine species are known to infect humans . Human infections are most frequent in Asia ( P . westermani , P . skrjabini , P . heterotremus , P . siamensis , P . miyazakiki ) , but they also occur in sub-Saharan Africa ( P . uterobilateralis , P . africanus ) , and in the Americas ( P . kellicotti , P . mexicanus ) [1] . Approximately 21 million people are infected with Paragonimus worms [2] , and some 293 million live in endemic areas where they are at risk of contracting the infection [3] . Paragonimus metacercariae enter the human host upon ingestion of raw or undercooked crustaceans . Metacercariae excyst , migrate out of the intestine , cross the diaphragm into the pleural space , and eventually invade the lungs where they mature and live for years in pulmonary cysts [1] . This results in a range of clinical symptoms , including cough , fever , weight loss , pleural effusion , chest pain , and bloody sputum [4] . These symptoms can be very similar to those seen in patients with tuberculosis , bacterial pneumonia , fungal infections , or lung cancer , so misdiagnosis is common [5]–[7] . For example , one study in the Philippines found P . westermani eggs rather than acid-fast bacilli in sputum samples from 26 of 160 ( 16% ) patients with suspected tuberculosis [5] . Even in the US , the median time between onset of symptoms and diagnosis of recent P . kellicotti infections was approximately 12 weeks ( range 3–38 weeks ) , and all of the patients were subjected to multiple , unnecessary medical interventions tailored to un-related diseases [8] . Once a proper diagnosis is made , parasites are easily cleared by a short course of the anthelmintic drug praziquantel , but infections can be fatal if left untreated [9] . Paragonimus infections are most often diagnosed by identification of parasite eggs in the stool or sputum ( reviewed in [1] ) . Unfortunately , migrating parasites are capable of causing disease weeks or months before eggs production commences . Egg detection is also insensitive due to temporal inconsistencies and requires knowledge and expertise that are not readily available in many clinical settings . Serological tests for P . westermani and P . kellicotti using native parasite antigens have been described , but these tests are impractical for widespread use because they require continued access to adult parasites [8] , [10] , [11] . Thus far , efforts to develop and implement practical , standardized molecular diagnostic tools have been hindered by a lack of information on the basic biology and genomics of Paragonimus species . According to the study outline presented in Figure 1 , we sequenced and annotated the transcriptome of adult P . kellicotti to better understand this parasite at a molecular level and to facilitate proteomic analyses of both the total worm homogenate and of immunogenic proteins purified using IgG from P . kellicotti patient sera . The resulting sequence data led to the identification of proteins that are promising candidates for the development of novel ( and much needed ) serodiagnostic tests for paragonimiasis . In addition , the annotated transcriptome of adult P . kellicotti provides a valuable resource for molecular biological and translational research on paragonimiasis and related food-borne trematode infections .
Wild crayfish ( genus Orconectes ) >3 cm in length were collected from small rivers in southern Missouri , USA . P . kellicotti metacercariae , identified by morphological examination , were isolated from the hearts of infected crayfish and introduced to Mongolian gerbils ( Meriones unguiculatus ) by intraperitoneal injection as previously described [12] . Gerbils were sacrificed 35–49 days post-infection , and egg-producing adult flukes were removed from lung cysts , rinsed in 1× phosphate buffered saline ( PBS ) , and stored at −80°C prior to use in experiments . Total RNA was isolated from two mature adult flukes using the PureLink RNA Mini Kit according to the manufacturer's microcentrifuge pestle protocol for animal tissues ( Ambion , Austin , TX ) , and DNase treated using the TURBO DNA-free Kit ( Ambion ) . cDNA was synthesized and sequenced as previously described [13] . Briefly , poly ( A ) RNA was selected from total RNA using the MicroPoly ( A ) Purist Kit ( Ambion ) and reverse transcribed using the Ovation RNA Amplification System V2 ( NuGEN Technologies , Inc . , San Carlos , CA ) . Paired-end , small fragment , Illumina libraries with insert sizes ranging from 180–380 bp were constructed and sequenced on an Illumina HiSeq2000 version 3 flow cell according to the manufacturer's recommended protocol ( Illumina Inc . , San Diego , CA ) . Raw reads were deposited in the NCBI sequence read archive under accession number SRX530756 ( NCBI BioProject Accession: PRJNA179523 ) . Raw reads were converted from bam to fastq format using Picard Tools' SamToFastq script ( http://picard . sourceforge . net ) . cDNA synthesis and Illumina sequencing adapters were trimmed using Flexbar [14] and Trimmomatic [15] , respectively . Trimmomatic was also used to perform sliding window quality trimming ( 5 bp window , average quality ≥20 ) and removal of reads less than 60 consecutive high quality bases and reads containing ambiguous base calls [15] . Reads with an average DUST score less than seven were removed using the filter_by_complexity script from the seq_crumbs package ( http://bioinf . comav . upv . es/seq_crumbs/ ) . Remaining reads were mapped against ribosomal RNA [16] , [17] and bacterial sequence databases [18] with Bowtie2 ( version 2 . 1 . 0 , default parameters , [19] ) and against the human genome ( hs37 ) and GenBank rodent database ( gbrod , downloaded April 24 , 2013 ) with Tophat2 ( version 2 . 0 . 8 , default parameters , [20] ) ; all matching reads and their mates were excluded from further analysis . The remaining high quality P . kellicotti originated reads were assembled using the Trinity de novo RNAseq assembler [21] with default parameters . Modules within the Trinity software package were used to estimate transcript abundance and remove transcripts representing <1% of the per-component expression level and <1 transcript per million [21] , [22] . The RNAseq reads used for the assembly were re-mapped to the high-confidence transcripts with Bowtie2 ( version 2 . 1 . 0 , default parameters , [19] ) and transcript breadth of coverage ( defined as the percent of covered bases over the length of the reference transcript ) was assessed using RefCov ( http://gmt . genome . wustl . edu/genome-shipit/gmt-refcov/current/ ) . Transcripts with <99% breadth of coverage with RNAseq reads were removed , resulting in the final transcript set . Assembly statistics at each phase of filtering are given in Table S1 . It is expected that the de novo assembly would over-estimate the number of transcripts and loci , so in-house PERL scripts were used to estimate fragmentation based on WU-BLAST alignments to protein coding sequences from closely related species as previously described [23] . Assembly fragmentation was calculated as the percentage of reference genes associated with multiple , non-overlapping BLAST hits . All assembled transcript isoforms were compared to known protein sequences by BLASTX [24] against the GenBank Non-Redundant protein database ( NR , downloaded April 15 , 2014 ) . Results were parsed to consider only top matches to non-overlapping regions of the query with e-value less than 1e-05 . Putative protein translations from the transcripts were predicted using Prot4EST [25] . Transmembrane domains and secretion peptides were predicted using Phobius [26] , [27] . Proteins were assigned to KEGG orthologus groups , biochemical pathways and pathway modules using KEGGscan [28] with KEGG release 68 . Associations with known InterPro domains and Gene Ontology ( GO ) classifications were inferred from predicted protein sequences using InterProScan [29]–[31] . Functional enrichment of GO terms was calculated using FUNC with an adjusted p-value cutoff of 0 . 01 [32] . For FUNC analysis , the target list included the longest isoform of a given locus that contained the feature of interest against the background of the longest isoforms of all loci including the target list . All transcripts , predicted proteins , and associated annotations are available at Trematode . net ( trematode . net/Paragonimus_kellicotti . html ) . Adult parasite antigen was prepared as previously described [12] . Briefly , eight adult parasites were homogenized on ice in RIPA buffer ( 10 mM Tris-HCl , pH 7 . 4 , 150 mM NaCl , 1% NP-40 , 0 . 2% sodium deoxycholate , 1 mM EDTA and 10 mM NaF ) using a 1 mL mini homogenizer ( GPE Scientific Limited , Leighton Buzzard , UK ) . The homogenate was centrifuged at 19 , 000×g for 15 minutes and the supernatant was collected . Protein concentration was measured using the Pierce BCA assay kit ( Thermo Scientific , Rockford , IL ) , and 500 µg was loaded onto GELFrEE 8100 fractionation system with an 8% cartridge ( Expedeon , San Diego , CA ) [33] , [34] . Eight molecular weight fractions were collected and the proteins were precipitated using a modified acetone-based method as previously described [35] . The pellets were solubilized in Tris buffer ( 100 mM Tris-HCl pH 8 . 5 ) containing 8M urea and the protein content was determined using the Advanced Protein Assay ( Cytoskeleton , Inc . , Denver , CO ) [ ( Fraction 1 ( F1 , lowest molecular weight ) , 35 µg; F2 , 176 µg; F3 , 126 µg; F4 , 83 µg; F5 , 71 µg; F6 , 67 µg; F7 , 76 µg; F8 , 40 µg ) ] . The quality of molecular weight fractionation was analyzed by SDS-PAGE; proteins were labeled with Sypro Ruby , and results were scanned using a Typhoon 9400 instrument . De-identified serum samples from P . kellicotti infected patients were obtained from Barnes Jewish Hospital in St . Louis , Missouri , the Centers for Disease Control and Prevention in Atlanta , GA , and Heartland Medical Center in St . Joseph , MO . Patients included in this study had reported ingestion of raw crayfish , exhibited symptoms consistent with paragonimiasis , tested positive for Paragonimus exposure using existing serological or parasitological diagnostic assays , and had no recent history of international travel . In all cases , sera were collected prior to treatment . Patient sera were tested for reactivity against adult P . kellicotti and P . westermani antigen by Western blot as previously described [10] . Serum samples from five strongly-reactive patients were pooled ( total volume 3 mL ) , and total IgG was precipitated using saturated ammonium sulfate ( Thermo Fisher Scientific , Pittsburg , PA ) , re-suspended in 1× phosphate buffered saline ( PBS ) , and desalted by dialysis against 4L 1× PBS for 2 hours at room temperature , against 4L 1× PBS 2 hours at 4°C , and against 4L 1× PBS overnight at 4°C . Two mL Pierce NHS-active agarose slurry ( Thermo Fisher Scientific ) was added to a 2 . 0 mL spin column ( Thermo Fisher Scientific ) , and rinsed with 2 . 0 mL water followed by 2 . 0 mL 1× PBS . Two mL of IgG precipitated from the paragonimiasis serum pool was added to the column and mixed for 2 hours at room temperature to couple IgG to the agarose . The column was washed once with 1× PBS , blocked with 1 . 0M ethanolamine pH 7 . 4 for 20 minutes at room temperature , and washed again with 1× PBS . Approximately 720 mg of adult P . kellicotti total antigen was added to the column and incubated overnight at 4°C . Column was washed with 1× PBS , and immune complexes were eluted with Pierce IgG elution buffer ( Thermo Scientific ) in eight 1 mL fractions . Fractions were neutralized with 50 µL 1 . 0M Tris , pH 9 . 0 , and 10 µL aliquots of each fraction were analyzed by Western blot as previously described using the pooled patient sera as the primary antibody [10] . The fraction with the highest concentration was precipitated using the 2D clean-up kit ( GE Healthcare , Buckinghamshire , UK ) and the pellet was solubilized in 20 µL 100 mM Tris-HCl , pH 8 . 5 with 8M urea to prepare peptides for mass spectrometry . The proteins that were eluted and denatured from the antibody coupled beads or from the GELFrEE protein fractions were reduced with 1 mM TCEP ( Pierce ) for 30 min , and alkylated with 20 mM Iodoacetamide ( Sigma ) at room temperature in the dark for 30 min . The reaction was quenched with 10 mM DTT ( Sigma ) for 15 min . Endoprotease Lys-C ( Sigma ) ( 5 µg ) was added and the samples were digested in a barocycler ( Pressure Biosciences ) [36] for 30 min at 37°C , followed by dilution to 2M urea with the Tris buffer , addition of trypsin ( Sigma ) and barocycler digestion for 30 min at 37°C . The digest was acidified to 5% formic acid and peptides were desalted in parallel on Glygen Nutips containing C4 and graphite carbon solid phase on a Beckman Biomek ( Biomek NXP ) , as previously described [37] . The eluted peptides were dried in a SpeedVac and dissolved in water/acetonitrile/formic acid ( 99%/1%/1% ) and transferred to autosampler vials ( SUNSRI Cat No . 200-046 ) for storage at −80°C or LC-MS analysis . Peptides for LC-MS from the GELFrEE fractionation were prepared as described above with the following modification . The endoprotease digests were acidified to 1% TFA , filtered through a 30K MWCO filter ( Sartorius VIVACON 500 ) . Peptides were desalted on a SepPak cartridge ( 50 mg/1cc ) ( Waters ) , dried in a SpeedVac and transferred into the autosampler vials for LC-MS analysis . A NanoLC 2D Plus System with a cHiPLC-Nanoflex and AS2 autosampler ( Eksigent , Dublin , CA ) was configured with two columns in parallel . One cHiPLC column ( ChromXP C18 ( 200 µm×15 cm; particle size 3 µm , 120 Å ) was used to inject calibrant solution ( β-galactosidase peptides ( 625 pmol/vial , part# 4333606 ) and another cHiPLC column was used for sample analysis . The calibrant solution ( 500 fmol ) was injected in solvent A ( water/formic acid/AcN , 98%/1%/1% ) . The samples were loaded in a volume of 10 µL at a flow rate of 0 . 8 µL/min followed by gradient elution of peptides at a flow rate of 800 nL/min . The calibrant solution was eluted with the following gradient conditions with solvent B ( water/formic acid/AcN , 1%/1%/98% ) :0 , 2%; 3 min , 2%; 73 min , 50%; 83 min , 80%; 86 min , 80%; 87 min 2%; 102 min , 2% . The digests from the immune-affinity purified samples were analyzed under the following gradient conditions ( time , percent solvent B ) : 0 , 2%; 3 min , 2%; 205 min , 35%; 215 min , 80%; 240 min , 2% . The digests from the GELFrEE fractionation were analyzed under the following gradient conditions ( time , percent solvent B ) : 0 , 2%; 5 min , 2%; 650 min , 35%; 695 min , 80%; 700 min , 2%; 720 min , 2% . Data acquisition was performed with a TripleTOF 5600+ mass spectrometer ( AB SCIEX , Concord , ON ) fitted with a Picoview Nanospray source ( PV400 ) ( New Objectives , Woburn , MA ) and a 10 µm Silica PicoTip emitter ( New Objectives , Woburn , MA ) . Data were acquired using an ion spray voltage of 2 . 9 kV , curtain gas of 20 PSI , nebulizer gas of 25 psi , and an interface heater temperature of 175°C . The MS was operated with a resolution of greater than or equal to 25 , 000fwhm for TOFMS scans . For data dependent acquisition , survey scans were acquired in 250 mS from which 100 product ion scans were selected for MS2 acquisition for a dwell time of 20 mS . Precursor charge state selection was set at +2 to +5 . The survey scan threshold was set to 100 counts per second . The total cycle time was fixed at 2 . 25 seconds . Four time bins were summed for each scan at a pulser frequency value of 15 . 4 kHz through monitoring of the 40 GHz multichannel TDC detector with four-anode/channel detection . A rolling collision energy was applied to all precursor ions for collision-induced dissociation using the equation , where the slope for all charges above 2+ is 0 . 0625 and the intercept is −3 , −5 and −6 for 2+ , 3+ , and 4+ , respectively . The raw LC-MS data ( * . wiff ) were converted to * . mzML format utilizing the AB SCIEX MS Data Converter v 1 . 3 ( AB SCIEX , Foster City , CA ) within PEAKS STUDIO 7 . 0 ( Bioinformatics Solutions Inc . , Waterloo , Canada ) . The resulting files were used for database searching by the PEAKS software using protein translations from the P . kellicotti transcriptome . The Ensembl Human protein database ( Homo_sapiens . GRCh37 . 72 ) was used to identify human background proteins in the sample matrix . The searches were conducted with trypsin cleavage specificity , allowing 3 missed cleavages , oxidation of Met and carbamidomethylation of Cys as variable and constant modifications , respectively . A parent ion tolerance of 25 ppm and a fragment ion tolerance of 100 milli-mass units were used . The MS2-based peptide identifications were validated within PEAKS software using a modified target decoy approach , decoy fusion , to estimate the FDR [38] . A 1% FDR for peptide spectral matches was used as the quality filter to identify peptides and associated proteins . MS data are available from Trematode . net ( trematode . net/Paragonimus_kellicotti . html ) and PeptideAtlas ( identifier PASS00555 ) . All animal work was performed in compliance with relevant US and international guidelines . Animal studies protocols were approved by the Washington University School of Medicine Animal Studies Committee ( Animal Welfare Assurance # A-3381-01 ) . The Animal Studies Committee complies with the United States Public Health Service Policy for Humane Care and Use of Laboratory Animals and other standards as required by the NIH Office of Laboratory Animal Welfare . The use of anonymized human sera was approved by the Washington University in St . Louis Institutional Review Board ( DHHS Federal Assurance #FWA00002284 ) under approval number 201102546 .
Prior to this study , a total of 911 GenBank sequences were available from the genus Paragonimus , only seven of which were from P . kellicotti . Therefore , it was necessary to sequence , assemble and analyze the transcriptome of P . kellicotti to enable further study ( Table 1 ) . Approximately 70 million paired-end reads were generated from an adult P . kellicotti cDNA library on the Illumina HiSeq platform . Following removal of low quality and contaminant reads , 40 million read pairs and 18 million unpaired orphan reads were assembled into 78 , 674 high-confidence transcript isoforms with an average length of 560 bp . These were further clustered into 54 , 622 distinct genetic loci , 21 . 5% of which are associated with more than one transcript isoform ( mean 3 . 0 transcript isoforms per alternatively spliced locus ) . We assume that the P . kellicotti genome contains a similar number of protein coding genes as other recently sequenced trematode genomes , which currently ranges from 10 , 852 in Schistosoma mansoni to 16 , 258 in Clonorchis sinenesis [39]–[43] . The discordance between the number of detected genetic loci and the expected number of genes is likely due to assembly fragmentation resulting in overestimation of the number of genes , a common problem seen in de novo transcriptome assemblies of short read data [44]–[46] . We calculated the fragmentation rate of our assembly at 25 . 8% using S . mansoni genes as a reference and at 31 . 4% using C . sinensis genes as a reference . The fragmentation rate is an estimate and it depends on the level of sequence conservation between the species of interest and species with available genome data; however , it is likely that at least 25 . 8–31 . 4% of all P . kellicotti genes represented in our assembly are split into two or more non-overlapping genetic loci . Assembled transcripts were compared to known proteins originating from other species . A total of 32 , 201 transcript isoforms from 20 , 102 loci shared a sequence similarity with an e-value cut-off of better than 1e-05 ( Table S2 ) . A majority of the matches were to sequences from C . sinensis followed by Schistosoma species . This is not surprising , as these were the only trematodes with sequenced genomes at the time this study was conducted . P . kellicotti sequences shared an average 61 . 3% sequence identity with corresponding C . sinensis sequences at the protein level . There were just 165 P . westermani sequences included in GenBank-NR at the time of this study , so only 125 transcripts from 67 genetic loci had a top BLASTX hit to a P . westermani protein . The sequence identity shared between P . kellicotti and P . westermani high-scoring segment pairs was 79 . 8% at the protein level . P . kellicotti and P . westermani are not considered to be close relatives within the genus Paragonimus [47]; however , the identified high level of sequence conservation may help facilitate the design of pan-Paragonimus serological assays . A total of 77 , 123 unique protein sequences were predicted from 54 , 616 of the detected genetic loci . Detailed annotations are available in Table S2 . Predicted proteins from 11 , 116 genetic loci were associated with a total of 4 , 407 unique InterPro protein domains and 1 , 234 unique GO terms . The number of genetic loci associated with each molecular function term was tallied , and the most abundantly represented terms were related to protein , ATP and nucleic acid binding . Similarly , the biological processes with the highest representation were protein phosphorylation , metabolic process , and oxidation-reduction process . In a comparison between three trematode species , a total of 312 conserved domains were unique to P . kellicotti , while 305 and 218 were unique to C . sinensis and S . mansoni , respectively ( Figure 2A ) . A majority of the domains present in each species were shared between all three species . Predicted proteins from 18 , 028 transcripts/11 , 599 genetic loci were associated with 6 , 854 unique KEGG orthologous groups . These were further binned into 336 unique biochemical pathways and 284 pathway modules . The KEGG orthologous groups represented in the adult of transcriptome of P . kellicotti were compared to those represented in the draft genomes of C . sinensis and S . mansoni ( Figure 2B ) . Altogether , 620 P . kellicotti KEGG orthologous groups ( KOs ) were absent from the other trematodes; these were binned into 255 pathways and 97 modules , most of which were very sparsely populated with the P . kellicotti-specific KOs . A careful analysis failed to identify any complete or nearly complete pathways present in P . kellicotti but absent in the other trematodes . The coverage of specific KEGG pathways can be visualized and compared to other trematodes using the TremaPath tool available at Trematode . net ( http://trematode . net/TN_frontpage . cgi ? navbar_selection=comparative_genomics&subnav_selection=tremapath ) . Secreted proteins have an important role in the life cycle of tissue-migrating parasite species like P . kellicotti , facilitating interactions with the host . These proteins are of practical interest as diagnostic , vaccine , or drug targets . Proteins related to 1 , 610 genetic loci were annotated as potentially secreted based on the presence of a classical signal peptide for secretion and absence of a predicted transmembrane domain ( Table S2 ) . Seven GO terms were found to be enriched among predicted secreted proteins , with the most highly enriched term being related to cysteine protease activity ( Table 2 ) . Proteases tend to be prevalent among trematode excretory-secretory products [48]–[51] , and various reports have described their role in migration through host tissues , nutrient uptake , and immune evasion [52]–[55] . Total parasite antigen was subjected to analysis by mass spectrometry to survey the worm proteome and subsequently to validate a subset of our assembled transcripts . A total of 244 , 048 spectra were matched to 25 , 405 database protein predictions that corresponded to 2 , 555 transcripts from 1 , 863 genetic loci ( Table S2 ) . The verified proteins encompass 1 , 626 InterPro protein domains , 586 GO terms , 1 , 925 KEGG orthologous groups from 307 pathways and 198 pathway modules . Furthermore , 63 transcripts from 48 genetic loci with no annotation ( i . e . , no significant BLAST hit in NR or KEGG , conserved protein domain , GO term , etc . ) were confirmed by the proteomic data . These sequences , thus far unique to P . kellicotti , might have otherwise been dismissed as low confidence transcripts due to the draft nature of the transcriptome assembly . However , proteomic evidence verified that these species-specific nucleotide sequences are translated and that they may have important biological functions in P . kellicotti . In order to obtain an estimate of abundance , identified proteins were ranked according to associated spectral counts . Given the draft nature of the transcriptome and the known issue of fragmentation , attempts were not made to correct for protein size , so follow up experiments would be required to assess abundance in a more robust and quantitative manner . The 25 proteins with highest spectral counts ( Table 3 ) included actins , myoglobins , chaperone proteins , and yolk ferritins , and these proteins may be abundant in the parasites . Oxygen binding proteins such as myoglobin are vital to parasite survival , as an exceptionally high affinity for their substrate allows the parasite to scavenge oxygen from host blood and tissues [56] . The high abundance of myoglobin proteins in our analysis may serve as an indication of the importance of aerobic respiration in P . kellicotti . Serodiagnostic assays based on worm homogenate have been shown to sensitively and specifically detect an immune response to P . westermani and P . kellicotti [8] , [10] , [11] . In these assays , total parasite protein antigens are analyzed by SDS PAGE gel electrophoresis , transferred to a membrane , and exposed to patient serum . Doublet bands appearing at 21/23 kDa and a more diffuse band at 34 kDa are indicative of exposure to Paragonimus species ( Figure 3 and [10] ) . However , the identity of these proteins was not known . An unusual cluster of cases of paragonimiasis ( caused by P . kellicotti ) occurred in recent years in the state of Missouri [8] , [57] , [58] . Since helminth infections are uncommon in Missouri , sera from these patients contain antibodies to Paragonimus antigens , but they are unlikely to contain antibodies to antigens of other helminths . These sera represented an excellent resource for our study . P . kellicotti proteins recognized by total IgG from some of these patients were enriched by immunoprecipitation using affinity beads . Eluate fractions were assessed by Western blot ( Figure 3 ) , and the strongest fraction was analyzed by mass spectrometry . A total of 2 , 406 spectra were matched to 1 , 443 proteins predicted from the transcriptome assembly that corresponded to 321 transcripts from 227 genetic loci ( Table S2 ) . Some 212 of these 227 loci were also detected in our analysis of the total worm proteome . Thus , the whole parasite proteome provided useful supplementary information to the immunoprecipitated proteins . The 25 most abundant proteins bound by patient IgG ( as approximated by spectral counts ) are listed in Table 4 . Most of the translations predicted from the transcriptome represent a fraction of the full length of the deduced protein . Therefore , it is challenging to determine with certainty which of these might represent the antigen present in the 21/23 kDa or 34 kDa bands . Nonetheless , several of the proteins on this list are of interest as potential serodiagnostic antigens . Five of the highly abundant immunoreactive proteins ( Table 4 ) , Pk00394_txpt2 , Pk45107_txpt2 , Pk48549_txpt1 , Pk24571_txpt1 , and Pk42039_txpt2 are putative cysteine proteases . Translations from three of these transcripts ( Pk00394_txpt2 , Pk48549_txpt1 , and Pk42039_txpt2 ) are predicted to have molecular weights in the range of 35–36 kDa , close in size to the diffuse ∼34 kDa antigen detected by serodiagnostic Western blots with total native parasite antigen ( Figure 3 ) . The predictions of 35–36 kDa are only estimates and may not represent the full length of the protein . However , the predicted molecular weights of top BLASTX hits of these proteins are in the same size range ( 36–37 kDa ) , and this indicates that the P . kellicotti sequences we have are complete or nearly so . Recombinant cysteine proteases have shown promise as serodiagnostic antigens for trematode infections [59]–[63] , and a previous study reported that partially purified cysteine proteases from P . westermani excretory-secretory products were superior for antibody diagnosis compared to whole worm antigen extracts [64] . Two of the most abundant proteins identified in the mass spectrometry analysis of our P . kellicotti immunoprecipitate , Pk00394_txpt2 and Pk48549_txpt1 , share 86% sequence identity at the amino acid level . These proteins are similar to cysteine proteases from other P . westermani and , to a lesser extent , helminths of other genera . By selecting a specific region from these cysteine proteases , it may be possible to develop an assay that discriminates between Paragonimus species and other helminths . A recombinant cysteine protease from P . westermani , rPwCP2 , has already shown promise as diagnostic antigen [62] , but this sequence ( gi:42516556 ) has no homolog in our P . kellicotti transcriptome . Thus , the cysteine proteases identified in our study may be more useful as a pan-Paragonimus diagnostic reagent than those previously described . Other proteins on our top-25 list ( Table 4 ) , such as the MF6p proteins and myoglobins , have not been considered as serodiagnostic antigens , but they are abundant excretory-secretory products of trematodes and merit further exploration . For example , Pk39524_txpt1 is annotated as a putative MF6p protein . Its top BLAST hit was recently characterized as a heme-binding protein and is a major antigen secreted by F . hepatica [65] . The P . kellicotti orthologue only shares 57% sequence identity with the F . hepatica protein , so cross-reactivity with antibodies in patients with fascioliasis should not be a major problem . Orthologs from other Paragonimus species have not yet been reported , so it is not possible to assess the potential utility of this protein as a pan-genus diagnostic reagent at this time . However , Pk34178_txpt1 , a putative myoglobin 1 , shares 90% sequence identity with an ortholog in P . westermani , but significantly less similarity with orthologs from other trematode species ( Figure 4 ) , strongly indicating that this candidate is worth further attention to examine its diagnostic utility . We undertook a systems biology approach to comprehensively study the adult transcriptome and proteome of P . kellicotti to improve understanding of the protein composition of the adult parasite and potential interactions between the parasite and its mammalian host . The transcriptome of adult P . kellicotti represents a major advance in the study of Paragonimus species . Transcriptomes provide powerful foundations for translational research in parasitology to develop improved diagnostic tests , treatments , and vaccines . In this study , transcriptome data was used together with immunoaffinity chromatography and mass spectrometry to efficiently identify candidate diagnostic antigens . Similar integrated approaches may be useful for identifying novel targets for drugs and vaccines . Finally , the data generated in this study ( transcriptome , proteome , and immunolome ) represent a valuable resource for the research community , and it will be especially helpful for annotating genomes of Paragonimus spp . as they become available .
|
Paragonimiasis is a food-borne trematode infection that people acquire when they eat raw or undercooked crustaceans . Disease symptoms ( including cough , fever , blood in sputum , etc . ) can be similar to those observed in patients with tuberculosis or bacterial pneumonia , frequently resulting in misdiagnosis . Although the infection is relatively easy to treat , diagnosis is complicated . Available diagnostic assays rely on total parasite homogenate to facilitate the detection of Paragonimus-specific antibodies in patients . Though these blot-based assays have shown high sensitivity and specificity , they are inconvenient because total parasite homogenate is not readily available . This study used next generation genomic and proteomic methods to identify transcripts and proteins expressed in adult Paragonimus flukes . We then used sera from patients infected with P . kellicotti to isolate immunoreactive proteins , and these were analyzed by mass spectrometry . The annotated transcriptome and the associated proteome of the antibody immune response represent a significant advance in research on Paragonimus . This information will be a valuable resource for further research on Paragonimus and paragonimiasis . Thus this project illustrates the potential power of employing systems biology for translational research in parasitology .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results/Discussion"
] |
[
"next-generation",
"sequencing",
"genome",
"expression",
"analysis",
"infectious",
"diseases",
"helminth",
"infections",
"medicine",
"and",
"health",
"sciences",
"genomics",
"foodborne",
"trematodiases",
"diagnostic",
"medicine",
"paragonimiasis",
"genome",
"analysis",
"serodiagnosis",
"transcriptome",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"parasitic",
"diseases",
"parasitology",
"foodborne",
"diseases"
] |
2014
|
Systems Biology Studies of Adult Paragonimus Lung Flukes Facilitate the Identification of Immunodominant Parasite Antigens
|
Epigenetic modifications influence gene expression and provide a unique mechanism for fine-tuning cellular differentiation and development in multicellular organisms . Here we report on the biological functions of UTX-1 , the Caenorhabditis elegans homologue of mammalian UTX , a histone demethylase specific for H3K27me2/3 . We demonstrate that utx-1 is an essential gene that is required for correct embryonic and postembryonic development . Consistent with its homology to UTX , UTX-1 regulates global levels of H3K27me2/3 in C . elegans . Surprisingly , we found that the catalytic activity is not required for the developmental function of this protein . Biochemical analysis identified UTX-1 as a component of a complex that includes SET-16 ( MLL ) , and genetic analysis indicates that the defects associated with loss of UTX-1 are likely mediated by compromised SET-16/UTX-1 complex activity . Taken together , these results demonstrate that UTX-1 is required for many aspects of nematode development; but , unexpectedly , this function is independent of its enzymatic activity .
The proper development of multicellular organisms requires strict regulation of cell-specific gene expression to ensure appropriate cell fate specification , cellular differentiation , and organogenesis . In addition to transcription factors , gene expression is controlled by chromatin organization , which is regulated by chromatin-remodelling factors and the post-translational modifications of histone proteins [1]–[3] . An important post-translational modification is the mono- ( me ) , di- ( me2 ) , or tri- ( me3 ) methylation of lysine residues ( K ) on the tail of histone 3 ( H3 ) . Specifically , the methylation of specific lysine residues plays a major role in the maintenance of active and silent gene expression states . The combination of H3 K4 , K36 , and K79 tri-methylation generally marks transcriptionally active regions , whereas H3 K9 and K27 tri-methylation marks regions of transcriptionally silenced genes [2] . The levels of methylation are modulated by the action of histone methyltransferases ( HMTs ) and histone demethylases ( HDMs ) . The largest group of histone demethylases contains a Jumonji C-domain ( JmjC ) that catalyzes the demethylation of specific lysine and arginine residues by an oxidative reaction requiring iron [Fe ( II ) ] and α-ketoglutarate ( αKG ) as cofactors [4] . There are 28 JmjC-containing proteins in humans , grouped in different families , and the majority of these are evolutionarily conserved [5] . The KDM6 subfamily ( UTX/UTY/JMJD3 ) was shown to catalyze the demethylation of H3K27me2/3 [6]–[11] , and the individual members were shown to regulate differentiation in several cellular systems [6] , [7] , [10] . In C . elegans , there are four KDM6 family members: jmjd-3 . 1 , jmjd-3 . 2 , jmjd-3 . 3 , closely related to JMJD3 , and utx-1 , the unique homologue of the human UTX/UTY . The functional role of these proteins in nematodes is not well defined . jmjd-3 . 1 has been reported to regulate somatic gonadal development [6] , while utx-1 has been implicated in vulva differentiation and aging [12]–[14] . In this report , we have analyzed the developmental functions of UTX-1 . We show that utx-1 plays a vital role during embryogenesis and acts in several aspects of nematode postembryonic development . Surprisingly , we found that the catalytic activity of UTX-1 is not of critical importance for UTX-1 function in development . Genetic and biochemical analyses indicate that UTX-1 acts through a SET-16 ( MLL ) /UTX-1 complex and that the primary role of UTX-1 resides in the regulation of the activity of this complex .
C . elegans D2021 . 1 encodes for a predicted protein of 134 kDa that has high homology and co-linearity with the mammalian UTX/UTY proteins ( Figure 1A ) ; thus we named this gene and its product utx-1 and UTX-1 , respectively . UTX-1 is expressed in most , if not all , nuclei of early and late stage embryos ( Figure 1B ) as well as during all of the larval stages and into adulthood ( Figure 1C ) , suggesting that UTX-1 could have a functional role throughout C . elegans development . To determine the biological function of UTX-1 two deletion mutant strains , utx-1 ( tm3136 ) and utx-1 ( tm3118 ) , were analyzed ( Figure 1A ) . The tm3136 allele is a 236 bp deletion that creates a premature stop codon , potentially encoding a truncated protein of only 28 amino acids , and very likely producing a null mutant . The tm3118 allele is an out–of-frame deletion of 547 bp . The truncated protein potentially retains the first 620 amino acids , but is lacking the JmjC domain and catalytic activity . The two alleles have similar phenotypes suggesting that they are both loss of function mutants . Homozygous utx-1 mutant worms that are derived from heterozygous mothers providing maternal UTX-1 , utx-1 ( m+/z− ) , are viable and reach adulthood . However , they produce only a few , mostly unviable , utx-1 ( m−/z− ) eggs ( Figure 1D and 1E ) , suggesting that UTX-1 is required for embryogenesis and that the lack of UTX-1 can be overcome by maternal contribution . Analysis of the dead embryos revealed that mutant utx-1 animals mainly arrested as late embryos ( Figure 1E ) . Dead L1 larvae , with misshapen bodies ( Figure 1E ) , were rarely observed ( 5% , n>200 ) . A putx-1::UTX-1::GFP ( UTX-1::GFP ) translational reporter as extra-chromosomal array was able to rescue the embryonic lethal phenotype observed in heterozygous utx-1 ( m−/z− ) progeny from mothers carrying either the tm3136 or tm3118 allele ( Figure 1D ) in several transgenic lines , leading us to conclude that UTX-1 is essential for embryogenesis and that the zygotic expression of UTX-1 is sufficient to restore embryonic viability . Indeed , progeny that did not receive the transgene from the mother , died as late stage embryos ( not shown ) or malformed L1 larvae ( Figure S1A ) , suggesting that UTX-1 is not required for very early embryogenesis . In agreement with this , analysis of epithelial junctions using an AJM-1::GFP translational reporter [15] suggests that the morphology of utx-1 ( m−/z− ) embryos , that did not inherit the transgene , was normal at early stages and progressively deteriorated throughout development ( Figure S1B ) . Analysis of markers for intestinal ( elt-2::GFP ) , muscular ( hlh-1::GFP and myo-2::GFP ) , and hypodermal ( dpy-7::GFP ) cells revealed that these cell lineages are correctly established in utx-1 ( tm3118 ) mutant worms ( Figures S2 , S3 , S4 , S5 ) . However , a progressive loss of myo-3::GFP transgene expression during embryogenesis was observed and little GFP signal was detected in L1 escapers ( Figure S6 ) , suggesting that defects in muscle function might account , at least in part , for the lethality of utx-1 null animals ( Figure 2B ) . To determine the function of UTX-1 at later developmental stages , we analyzed both the utx-1 ( m+/z− ) mutant worms , in which the zygotic contribution of UTX-1 is lost , and wild-type worms in which utx-1 expression was downregulated by feeding RNA interference ( RNAi ) . RNA interference , with constructs targeted to three different regions of utx-1 ( Figure 1A ) , resulted in an approximately 60% reduction of utx-1 mRNA in F1 progeny and in a significant reduction of UTX-1 protein expression ( Figure S7A ) . In agreement with the phenotype of utx-1 ( m+/z− ) mutant animals , the utx-1 ( RNAi ) F1 animals had reduced fertility ( Figure S7B ) . Furthermore , variable defects , often located posteriorly , were observed in about 40% of the utx-1 ( RNAi ) worms ( Figure 2A and 2C ) . The posterior defects observed in animals treated with utx-1 ( RNAi ) were very similar to the defects observed in utx-1 ( m−/z− ) dead larvae ( Figure 2A , 2C and Figure S8 ) . This demonstrates that RNA interference can be used to efficiently analyze the postembryonic roles of UTX-1 , and that the posterior phenotype in utx-1 ( m−/z− ) is due to the loss of utx-1 . Importantly , transgenic expression of wild-type utx-1 fully rescued the posterior defects in larvae . The reduced fertility observed in utx-1 ( m+/z− ) animals might be due to a regulatory role for UTX-1 in either somatic gonad or germline development . Homozygous mutant utx-1 ( m+/z− ) animals from heterozygous animals generally develop germlines with correct proliferation and differentiation patterns ( not shown ) , as demonstrated by the fact that oocytes are formed ( Figure 2B ) and by an ability to lay a few dead embryos , suggesting that the sterility is not related to a germline defect . However , animals lacking UTX-1 activity have defects in gonad migration and oocyte organization . The shape of the gonad is dictated by the coordinated migration of two distal tip cells ( DTCs ) , which are part of the somatic gonad structure and move away from the gonad primordium during postembryonic development , leading to two consecutive turns forming the U-shaped gonad arms observed in adult animals . Using transgenic animals carrying a distal tip cell marker , lag-2::GFP [16] , we observed aberrant gonadal migration in 42% ( n = 176 ) of the utx-1 ( RNAi ) animals . Morphological analysis by DIC of utx-1 ( RNAi ) animals further confirmed that 48% ( n = 215 ) of the animals showed a failure to turn or abnormal turning of at least one gonad arm ( Figure 2B and 2C ) , and these animals often ( 41% , n = 137 ) developed misshapen gonads , with an enlargement of the proximal end of the gonad arms and a misorganization of oocytes ( Figure 2B ) . These gonad phenotypes were also identified in utx-1 ( m+/z− ) mutant animals ( Figure 2B , 2C and Figure S9 ) , and they were efficiently rescued by the UTX-1::GFP transgene , reinforcing that these aberrations are caused by the loss of utx-1 ( Figure 2C ) . The fact that the transgenic expression of wild-type utx-1 is able to rescue the sterility and the gonadal phenotypes suggests that utx-1 has a role in the somatic gonad rather than in the germline , where transgenes are normally silenced . Consistent with this , GFP-tagged UTX-1 is expressed in the distal tip cells during migration ( Figure 1C ) and other tissues of the somatic gonad , such as the sheath cells and the spermatheca ( not shown ) and not in the germline . The aberrant migration and oocyte organization defects are similar those we reported for jmjd-3 . 1 loss-of-function mutants , which encodes one of the C . elegans homologues of the JMJD3 family [6] . To determine if there is a link between these two observations , we tested if UTX-1 affected the expression of jmjd-3 . 1 by performing quantitative PCR on utx-1 ( RNAi ) animals . As shown in Figure 2D , utx-1 ( RNAi ) animals have reduced levels of jmjd-3 . 1 , suggesting that UTX-1 may , directly or indirectly , regulate jmjd-3 . 1 expression . Additionally , an enhancement of the phenotype was not observed when utx-1 was reduced in a jmjd-3 . 1 mutant genetic background ( see below ) , suggesting that both genes are acting in the same genetic pathway to regulate somatic gonadal development . UTX-1 belongs to the KDM6 family , of which members have been shown to catalyze the demethylation of H3K27me3 and H3K27me2 [17] . Several observations indicate that this role is conserved throughout the C . elegans life cycle . First , loss of the zygotic and maternal contributions of utx-1 results in increased global levels of H3K27me2/3 at the embryonic stage ( Figure 3A ) . Second , reduction of UTX-1 by RNA interference results in a significant increase of H3K27me3 levels at different larval stages ( data not shown , [12] , [18] ) . Third , exogenous expression of wild-type UTX-1 in utx-1 null animals restores H3K27me3 to wild-type level ( Figure 3A ) . Fourth , over-expression of UTX-1 in wild-type animals results in a significant reduction of global H3K27me3 levels ( Figure 3B ) . Finally , the decreased level of H3K27me3 observed in N2 worms overexpressing UTX-1 ( Figure 3B ) is well correlated with the degree of UTX-1::GFP overexpression , as shown in Figure S10 . Next , we tested if the catalytic activity of UTX-1 is responsible for the phenotype observed in utx-1 null mutants and utx-1 ( RNAi ) animals . To this end , we expressed in utx-1 mutants a GFP-tagged mutated form of the UTX-1 protein ( for simplicity called UTX-1DD::GFP , DD = Demethylase Dead ) , carrying mutations in two of the three conserved amino acids in the iron-binding motif ( HXD/EXnH ) of the JmjC-domain ( indicated by asterisks in Figure 3C ) . Several reports have shown that these amino acids are required for iron binding and thus for the catalytic activity of all JmjC-domain containing demethylases characterized so far [6]–[11] , [19]–[21] . All UTX-1::DD::GFP transgenic lines generated ( 8/8 ) showed expression at levels similar to wild-type UTX-1 ( Figure 3C ) and were fertile and able to produce viable progeny ( Figure 1D ) . Importantly , re-expression of catalytically inactive UTX-1 did not restore the wild-type level of H3K27me3 in utx-1 null animals ( Figure 3A ) and did not influence the H3K27me3 level when overexpressed in wild-type animals ( Figure 3B ) , thus confirming that the amino acids substitutions affected UTX-1 enzymatic activity . This unexpected result strongly indicates that the demethylase activity of UTX-1 is not important for either embryonic development or animal viability . Subsequently , we tested if the other observable phenotypes were dependent on UTX-1 enzymatic activity . Tail and gonadal defects were also efficiently rescued ( Figure 2C ) in 50% ( 4/8 ) of the transgenic lines , indicating that UTX-1 , but not its catalytic activity , is required for correct posterior and gonadal development . jmjd-3 . 1 , jmjd-3 . 2 , and jmjd-3 . 3 ( Figure 4A ) are C . elegans KDM6 family members closely related to human JMJD3 . Animals carrying mutations in one of these genes are viable , fertile ( not shown ) , and do not show up-regulated levels of H3K27me3 by western blot analysis ( Figure 4B and Figure S11B ) . However , triple mutant worms carrying deletions in all three JMJD3-like genes showed increased global levels of H3K27me3 ( Figure 4B and Figure S11B ) , suggesting that these proteins are H3K27me3 demethylases and might act redundantly with UTX-1 . Several lines of evidences indicate that the JMJD3-like genes do not function redundantly with UTX-1 . Analysis of the transcriptional expression levels of the JMJD3-like genes in wild-type worms indicated that only jmjd-3 . 1 is expressed at levels comparable to utx-1 , while jmjd-3 . 2 and jmjd-3 . 3 are only weakly expressed , in particular during larval stages ( Figure S11A ) . Furthermore , the transcriptional expression pattern of the JMJD3-like genes appeared generally restricted to specific tissues or , as in the case for jmjd-3 . 2 , even to few cells ( Figure S11C ) ; this is in contrast to the broad expression pattern of UTX-1 . In addition , the triple mutant lacking the JMJD3-like genes is viable and fertile , with no defects in the posterior region of the body ( Figure 4C ) and with only minor gonadal defects ( Figure 4C ) , most likely due to the absence of jmjd-3 . 1 . Importantly , the down-regulation of utx-1 by RNA interference in the triple mutant genetic background did not exacerbate the posterior or the gonadal defects associated with utx-1 reduction in wild-type animals ( Figure 4C ) . Taken together , these results imply that the members of the KDM6 class do not act redundantly . However , in light of the unexpected results obtained with the catalytically inactive UTX-1 mutant , it is important to take into consideration the possibility that JMJD3-like genes could , nevertheless , compensate for the lack of UTX-1 activity in utx-1 mutant worms expressing the catalytically inactive form of UTX-1 . In this case , we would expect that the loss or reduction of other H3K27me3 demethylases in the utx-1 mutant rescued with the catalytically inactive UTX-1 would result in utx-1-specific abnormalities ( posterior defects and aberrant gonadal migration ) . To test this hypothesis , we generated a triple mutant jmjd-3 . 2; jmjd-3 . 3;utx-1+UTX-1DD::GFP in which the fourth member of the KDM6 family , jmjd-3 . 1 , was down-regulated by RNA interference . In this genetic background , no posterior defects were observed and the degree of gonadal defects was similar to those observed in wild-type animals under the same conditions ( Figure 4D ) . Furthermore , quantitative PCR showed no increased expression levels of the JMJD3-like genes in the rescued transgenic line utx-1+UTX-1DD::GFP compared to utx-1+UTX-1::GFP ( Figure 4E ) . These results , together with the fact H3K27me3 levels are still up-regulated in utx-1 expressing the catalytically inactive form of UTX-1 ( Figure 3A ) , strongly indicate that the JMJD3-like proteins do not compensate for the lack of UTX-1 catalytic activity and that the catalytic activity of UTX-1 is not required for proper development . The mammalian UTX is part of the MLL3/MLL4 H3K4me3 methyltransferase complex [22]–[24] that also includes the specific component PTIP , and WDR5 , ASH2L , and RbBP5 as core components , which are also shared by other complexes [25] . The high conservation of these proteins in nematodes ( WDR5/tag-125/wdr-5 . 1 , ASH2L/ash-2 , RbBP5/F21H12 . 1/rbbp-5 , MLL3-4/set-16 , UTX/utx-1 , and PTIP/pis-1 ) , suggests that a similar complex could also exist in C . elegans . To test if an MLL3-4/UTX-like complex ( SET-16/UTX-1 ) is present in C . elegans , we purified GFP-tagged UTX-1 and associated proteins from a mixed population of transgenic animals , enriched with embryos ( Figure 5A ) . The identities of the interacting proteins were determined by mass spectrometry and are listed in the Table S1 . As a control , N2 lysates were subject to the same procedure and the recovered proteins ( listed in Table S2 ) were considered contaminants and used to confirm the specificity of the identified interacting proteins . Strikingly , all of homologous components of the mammalian MLL3/4 complex were identified as UTX-1 partners in C . elegans ( Figure 5B ) . As further verification , we utilized a transgenic line carrying HA-tagged WDR-5 . 1 [26] , the most prominent WDR5-like protein recovered by mass spectrometry , in which we expressed UTX-1::GFP . As shown in Figure 5C , in lysates derived from embryos , both UTX-1::GFP and endogenous UTX-1 were found associated with WDR5 . 1 , further supporting the existence of a SET-16/UTX-1 complex in C . elegans . Importantly , the catalytically inactive mutant UTX-1DD::GFP was also recovered by WDR-5 . 1 immunoprecipitation ( Figure 5C ) . Gel filtration analysis of lysates from transgenic lines carrying either the wild-type or the catalytically inactive forms of UTX-1 further confirmed that both UTX-1 and UTX-1DD are engaged in large complexes ( Figure 5D ) , further supporting that a functional JmjC domain is not required for the association with the complex . We then verified the functional correlation of the SET-16/UTX-1 complex components by testing if their loss or downregulation could result in phenotypes similar to those observed in the utx-1 mutant . Loss of set-16 results in embryonic and early larval lethality [27] . The analysis of set-16 ( n4526 ) young larvae revealed the presence of posterior defects similar to those identified in utx-1 null animals ( Figure 6A and Figure S8 ) , and set-16 ( RNAi ) animals that escaped embryonic and early larval lethality , often had abnormal gonad migration and enlargement ( Figure 6A , 6B and Figure S9 ) , which phenocopied the effect of the loss of utx-1 . Similarly , in pis-1 ( ok3720 ) mutants and pis-1 ( RNAi ) animals , posterior and gonadal defects were observed , although at a lower degree ( Figure 6A and 6B , Figures S8 and S9 ) . RNA interference of the core components of the complex ( F21H12 . 1 , wdr-5 . 1 , and ash-2 ) also resulted in posterior and gonadal defects similar to the ones observed in utx-1 mutants ( Figure 6C and Figure S9 ) . It should be noted , that enlargement of the proximal gonad was never observed after the reduction by RNAi of F21H12 . 1 and ash-2 and was rarely observed in wdr-5 ( RNAi ) animals ( Figure S9 ) . We then tested the effects of simultaneously downregulating specific components of the complex by RNAi . As shown in Figure 6B , the concurrent knockdown of utx-1 and set-16 or pis-1 did not enhance the phenotypes; similar results were obtained with concomitant silencing of pis-1 and set-16 . The high degree of phenotypic similarity and the absence of redundancy are evidence that these genes are acting in the same genetic pathways to regulate posterior patterning and somatic gonadal development . Along the same line , qPCR analysis revealed that set-16 downregulation by RNA interference results in a reduction of jmjd-3 . 1 mRNA ( about 60% decrease compared to control RNAi , data not shown ) , further supporting the notion that UTX-1 and SET-16 act in the same complex . Since the catalytic activity of UTX-1 is not necessary to rescue the developmental defects observed both in utx-1 mutants and in animals in which different factors of the complex were lost or down-regulated , we hypothesized that UTX-1 might regulate the expression of other components of the complex . In support of this , we found that the levels of set-16 mRNA were reduced in utx-1 ( RNAi ) animals ( Figure 6D ) . Interestingly , downregulation of set-16 also results in decreased expression of utx-1 mRNA and protein ( Figure 6D and 6E ) , suggesting an interdependent regulation of , at least , these two members of the complex . Taken together the data demonstrate that the SET-16/UTX-1 complex is present in C . elegans , and it is required for development . That the loss or downregulation of single components of the complex results in similar phenotypes as those observed in utx-1 null mutants , indicates that each component is required for the complex to function normally and that the defects associated with the loss of UTX-1 are likely the result of compromised SET-16/UTX-1 complex activity .
We have demonstrated that C . elegans UTX-1 is an H3K27me2/3 demethylase that is essential for development during embryonic and larval stages of the nematode , independently of its demethylase activity . Animals lacking the maternal and zygotic contribution of UTX-1 arrest during late embryogenesis . Although , analyses of reporter genes revealed no major defects in lineage specifications , a reduction of myo-3::GFP , expression , but not hlh-1::GFP , was observed in utx-1 mutant animals , suggesting that utx-1 might regulate genes involved in muscle function . In agreement , mammalian UTX has been implicated in terminal differentiation of muscle cells [28] . The maternal contribution of UTX-1 allows utx-1 ( m+/z− ) worms to reach adulthood , but defects arise at different stages of development , including abnormal gonad migration and oocyte misorganization . This latter phenotype could explain , at least in part , the reduced fertility of utx-1 ( m+/z− ) animals . We have previously shown that proper gonad migration partly depends on another H3K27me3 demethylase , jmjd-3 . 1 [6] . The expression level of jmjd-3 . 1 is significantly reduced in utx-1 ( RNAi ) animals . However , it should be noted that the loss of utx-1 leads to a more severe phenotype than the loss of jmjd-3 . 1 , which only influences gonadal processes at high temperature and moderately reduces fertility . These results suggest that utx-1 , in addition to jmjd-3 . 1 , modulate additional genes required for establishing the correct developmental program of gonads . While utx-1 represents the unique UTX/UTY homologue , C . elegans has three other genes with homology to the single mammalian JMJD3 gene ( jmjd-3 . 1 , jmjd-3 . 2 and jmjd-3 . 3 ) . We generated mutant animals carrying mutations in all three JMJD3-like genes and , unexpectedly , we did not detect any additional phenotypes in the triple mutants , other than the phenotypes already reported for jmjd-3 . 1 [6] . While it is possible that residual gene function remains in these mutants , the global level of H3K27me3 was significantly increased in the triple knockout worms , whereas no increase was observed in the jmjd-3 . 1 mutant strain alone ( [6]; Figure 4B ) . This data suggests that the JMJD3-like demethylases might regulate the expression of restricted sets of genes or that they have overlapping functions . Our analysis of the global levels of H3K27me2/3 also suggests that UTX-1 is the most important demethylase for the removal of the H3K27me3 mark among the members of the KDM6 family . Accordingly , the loss of utx-1 results in sterility ( in m+/z− worms ) and in embryonic lethality ( in m−/z− worms ) while animals lacking the three JMJD3 homologues are fertile and viable . This result indicates that utx-1 plays unique and essential roles during embryonic and postembryonic development and suggests that the JMJD3-like proteins , like the human homologues [10] , [29] , are mainly required for regulating cellular responses under particular conditions , such as stress or aging . Strikingly , we found that the catalytic activity of C . elegans UTX-1 is not required for the function of the protein in the developmental processes analyzed . This is at odds with a previous report describing the role of utx1 genes in D . rerio , in which human wild-type , but not the catalytically inactive mutant , partially rescued the defects in UTX morphant animals [9] . We do not know if this apparent dissimilarity is due to an organismal difference , as suggested by the fact that C . elegans UTX-1 does not regulate HOX genes ( data not shown ) as it does in zebrafish [9] and that zebrafish has two UTX homologues , or to the different experimental approaches . Interestingly , recent results also suggest a catalytic-independent role for human JMJD3 and UTX in chromatin remodeling in a subset of T-box target genes [30] . Quantitative PCR and analysis of reporter genes failed , however , to identify any regulation of selected C . elegans T-box genes by UTX-1 ( not shown ) . The demonstration that UTX , which mediates H3K27me2/3 demethylase activity , is part of the MLL3/4 complex , which also has H3K4 methyltransferase activity [6] , [7] , suggests a model in which the coordinated removal of repressive marks ( H3K27me3 ) and the deposition of activating marks ( H3K4me3 ) fine-tune transcription during differentiation . We have shown that a similar complex is present in C . elegans , and that it is required to achieve proper development . Indeed , loss or reduction of each component of the complex results in phenotypes similar to those we observed in utx-1 mutants . The lack of synergistic effects in double RNAi experiments further supports the notion that the components of the complex act in the same pathway ( s ) to regulate posterior body and somatic gonad development . Surprisingly , utx-1 phenotypes are rescued by catalytically inactive UTX-1 . The catalytically inactive mutant binds WDR-5 . 1 similarly to the wild-type protein , and it was identified in gel filtration experiments in a large complex , similarly to its wild-type counterpart . WDR-5 . 1 is also a component of other complexes and we cannot exclude at this time that the UTX-1/WDR-5 . 1 interaction might take place in the context of another complex . However , the components of other complexes with which WDR-5 . 1 is involved have , thus far , not been recovered by our mass spectrometry analysis . For example we did not identify the known WDR-5 . 1 binding partner SET-2 ( the main H3K4me3 methyltransferase in C . elegans ) [26] , [31] , [32] , suggesting that UTX-1 is specifically recruited in the SET-16 ( MLL ) -like complex . Taken together these results strongly suggest that UTX-1 acts through a SET-16/UTX-1 complex and indicate that the primary role of UTX-1 in C . elegans development is independent of the demethylase activity , possibly through the regulation of expression of the complex components . This is suggested by our results showing that UTX-1 is , at least , required for the proper expression of set-16 , and that SET-16 is required for the expression of utx-1 , suggesting a positive feed forward mechanism for retaining the activity of the SET-16/UTX-1 complex . It is possible that there are additional functions for UTX-1; UTX-1 may be required for targeting the complex to specific genomic regions or it might play a role in the stability of the complex . To correctly address these possibilities , chromatin immunoprecipitation and mass spectrometry analysis must be performed in the context of utx-1 null mutants . Unfortunately , these experiments are currently unfeasible since the utx-1 mutant is unviable . It should be mentioned , however , that downregulation of the human UTX does not interfere with MLL complex formation ( Agger K . , Helin K . , unpublished data ) , at least in mammals . Finally , we do not know if UTX-1 works exclusively in association with the SET-16 complex or if it has additional roles as a single protein or in association with other complexes . The results obtained by mass spectrometry analysis suggest that this latter hypothesis might be correct . UTX-1 immunoprecipitates with other proteins involved in distinct chromatin complexes , such as HDA-1 and LIN-53 , components of the NuRD complex [33] , and MIX-1 , which functions in the dosage compensation complex ( DCC ) [34] . While these interactions await further validation , it is worth noting that elements of the NuRD complex have been involved in vulva formation [33] , a postembryonic event in which both UTX-1 and SET-16 have been implicated ( [12] and data not shown ) , and that the DCC has been recently shown to interact with ASH-2 , a member of the complex that we describe here . Therefore , it is conceivable that UTX-1 works in diverse chromatin complexes to accomplish functions required at different stages of development or under specific conditions . A MLL complex-independent role for UTX is also supported by the fact that a substantial amount of mammalian UTX is bound to promoter regions depleted of H3K4 methyl marks [35] . We did not detect reduced global levels of H3K4me3 in utx-1 mutant animals ( data not shown ) , which could be expected if UTX-1 regulates the function of a complex having H3K4 methyltransferase activity . This is in agreement with previously reported result in mammals and C . elegans [7] , [12] and it is consistent with the fact that inactivation or downregulation of set-16 only results in a minor , if any , reduction in global levels of H3K4me3 [12] , [35] , [36] . Indeed , similar to mammals , H3K4me3 deposition in C . elegans is mainly regulated by the other H3K4me3 methyltransferase , set-2 [26] , [36] . This observation suggests that the SET-16/UTX-1 complex regulates the mark deposition only for a subset of genes , and , consequently , complex impairment does not impact the global levels H3K4me3 . Our analysis failed to uncover a role for UTX-1 catalytic activity during development , and a major question is therefore whether this activity is required for any biological function in C . elegans . Since this work focused on the role of UTX-1 during development , the catalytic activity might be required for other processes that are not implicated in developmental programs and are dispensable for viability . Indeed , recent reports implicate the catalytic activity of UTX-1 in aging [13] , [18] . Moreover , UTX-1 activity could act during germline formation to counteract the well-established role of the PRC2/MES complex [37]–[43] . However , we have thus far not been able to establish a function of UTX-1 during germline formation neither alone nor in synthetic interaction with components of the MES complex ( data not shown ) . In summary , we have shown that UTX-1 plays an essential role in several developmental processes in C . elegans . Surprisingly , the catalytic activity is dispensable for proper development , and our data suggest that UTX-1 acts , instead , through a SET-16/UTX-1 complex . Future studies will be directed at identifying the specific target genes regulated by the complex and the possible role that UTX-1 might play in the stability of the complex and in its recruitment to the target genes .
C . elegans strains were cultured using standard methods [44] . Strains used were as follows: wild-type Bristol ( N2 ) , utx-1 ( tm3118 ) X , utx-1 ( tm3136 ) X , jmjd-3 . 1 ( gk384 ) X , jmjd-3 . 2 ( tm3121 ) X , jmjd-3 . 3 ( tm3197 ) X , JK2049 qls19 V , set-16 ( n4526 ) III , pis-1 ( ok3720 ) IV , AZ217 ( myo-2::GFP ) , MS438 ( elt-2::GFP ) , GS3798 ( dpy-7::YFP ) , OP64 ( hlh-1::GFP ) , PS3729 ( ajm-1::GFP ) . The strain wdr-5 . 1/tag-125::HA and OE4201 ( myo-3::GFP ) were generous gifts from Francesca Palladino and Thomas Bürglin , respectively . Transgenic animals with specific genetic backgrounds were generated by standard crossing procedure . The C . elegans utx-1 sequence is located on chromosome X and the transcript encompasses 14 exons coding for a predicted protein of 1168 amino acids . The ATG of the gene is located at position 13888 bp of the D2021 cosmid ( U23513 ) and a TAG terminator codon at position 19549 bp . Two alleles of utx-1 were identified at the National BioResource Project ( NBRP ) , Japan . Both alleles were backcrossed three times with N2 before the phenotypic analysis and maintained in culture as heterozygotes . The tm3136 allele lacks 236 bp and the deletion is found at position 13920–14155 bp of the Genbank entry U23513 . This deletion creates a premature stop codon and the deleted gene could potentially encode for a truncated protein of 28 amino acids . The tm3118 is an out–of-frame deletion of 547 bp situated at position 17361–17907 bp of the Genbank entry U23513 . The truncated protein potentially retains the first 620 amino acids and lacks the JmjC domain . Phenotypic analyses of utx-1 mutant animals ( tm3136 and tm3118 ) were done in blind , before genotyping . jmjd-3 . 2 ( tm3121 ) X and jmjd-3 . 3 ( tm3197 ) X were backcrossed three times before analysis and their deletions , described in Wormbase , were confirmed by sequencing . KDM6 members are all positioned on the X-chromosome , with utx-1 located very closely to jmjd-3 . 1 , thus precluding the generation of a quadruple mutant lacking all the four H3K27me3 demethylases . The triple mutant with deletions in jmjd-3 . 1 ( gk384 ) X , jmjd-3 . 3 ( tm3197 ) X and jmjd-3 . 2 ( tm3121 ) X was generated using standard crossing methods . The triple mutant jmjd-3 . 2 ( tm3121 ) ;jmjd-3 . 3 ( tm3197 ) ;utx-1 ( tm3118 ) +UTX-1DD::GFP ( expressing UTX-1DD as an extrachromosomal array ) was generated using standard crossing methods . Single fourth-stage ( L4 ) larvae were plated in agar plates with OP50 bacteria and moved to a new plate every 24 h . Viable progeny were counted every day , for 4 days at 25°C . The average number of progeny produced by a single animal is reported . RNAi was performed by feeding and carried out as described previously [45] . For UTX-1 , a clone ( X-4I10 ) containing the region from 18167 bp to 19214 bp of the GenBank entry U23513 ( c in Figure 1A ) , was obtained from the C . elegans RNAi feeding library ( J . Ahringer's laboratory , Wellcome Trust/Cancer Research UK Gurdon Institute , University of Cambridge , Cambridge , UK ) . Two other clones ( a and b in Figure 1A ) , spanning the regions 14109–14331 bp and 17512–18014 bp of the GenBank entry U23513 , respectively , were constructed by PCR . We generated RNAi clones for set-16 , ash-2 , pis-1 , tag-125 , and F21H12 . 1 by amplifying cDNA fragments ( approximately 500 pb ) before cloning in L4440 plasmid using EcoRI restriction sites ( all primer sequences available upon request ) . Eggs , prepared by hypochlorite treatment , were added onto RNAi bacteria-seeded NMG plates and cultivated at 25°C . Control animals were fed with bacteria carrying the control vector ( L4440 ) . Generally , F1 progeny was scored for phenotypes . Total RNA was isolated from eggs using TRIzol reagent ( Invitrogen ) and RNAeasy Minikit ( Qiagen ) . cDNA was synthesized using reagents from the TaqMan Reverse Transcription kit ( Applied Biosystems ) . qPCR was performed using SYBR Green 2× PCR Master mix ( Applied Biosystems ) in an ABI Prism 7300 Real Time PCR system ( Applied Biosystems ) . The measures were normalized to ribosomal protein ( rpl-26 ) RNA levels . All reactions were performed in triplicate , in at least three independent experiments . All primer sequences are available upon request . For mutant strains , total protein extracts were prepared from eggs obtained by hypochlorite treatment of adults grown on OP50 at 25°C . For RNAi-treated animals , extracts were prepared from eggs obtained by hypochlorite treatment of adults grown on HT115 containing either the empty feeding vector , or specific RNAi . Protein concentration was estimated using the modified micro-Lowry assay and equal amounts of protein were loaded . The following antibodies were used: polyclonal anti-H3 ( Abcam 1791 , lot GR9204-1 ) 1∶30000; polyclonal anti-H3K27me1 ( Upstate 07-448 , lot DAM1598790 ) 1∶5000 , polyclonal anti-H3K27me2 ( Abcam 24684 , lot 956943 ) 1∶2000; polyclonal anti-H3K27me3 ( Upstate 07-449 , lot 701050758 ) 1∶2000; monoclonal anti-actin ( Chemicon International MAB1501 ) 1∶10000; peroxidase-labeled anti-rabbit and anti-mouse secondary antibodies ( Vector ) . The specificity of H3K27 antibodies has been tested as shown in Figure S12 . Polyclonal C . elegans UTX-1 antibodies were obtained through the Eurogentec polyclonal antibody production service . To generate a specific UTX-1 antiserum , rabbits were immunized with two UTX-1 peptides ( MDESEPLPEERHPGNC and SYRRSYKDDANRLDHC ) . Antibodies were purified using affinity columns coupled with the same peptides and used at 1∶500 dilution . The antibody recognizes in the lysate of wild-type animals a specific band of the predicted size of 134 kDa , absent in the lysates obtained from utx-1 mutant alleles ( Figure S12D ) . Western blots were quantified using ImageJ program ( National Institutes of Health ) . For the UTX-1::GFP construct , a 6956-bp fragment of utx-1 including 1290 bp of promoter region and the entire coding region was PCR-amplified from N2 genomic DNA . The resulting fragment was inserted in the multiple cloning site of the pPD95 . 75 vector ( Fire lab ) . For the UTX-1DD::GFP construct , the UTX-1::GFP construct was mutated using the QuikChange Site-Directed Mutagenesis Kit ( Stratagene ) . Specifically , the DNA sequence was mutated so that the histidine at position 914 ( H914 ) and the aspartic acid at position 916 ( D916 ) were changed to alanine . The DNA sequences of both constructs were verified by sequencing . To obtain lines carrying extra-chromosomal arrays , 20 ng/µl of UTX-1::GFP and UTX-1 DD::GFP constructs were each co-injected with 100 ng/µl pRF4 ( rol-6 ( su1006 ) ) or ttx-3p::RFP in wild-type N2 worms ( N2+UTX-1::GFP and N2+UTX-1DD::GFP ) . Transgenic lines in utx-1 ( tm3136 and tm3118 ) genetic backgrounds ( utx-1+UTX-1::GFP and utx-1+UTX-1DD::GFP ) were generated by crossing . Fluorescence microscope and DIC pictures were acquired using an Axiovert 135 , Carl Zeiss , Inc . with a 63× Plan Apochrome objective with a NA of 1 . 4 in immersion oil and a 40× Plan NEOFLUAR with a NA of 0 . 75 , respectively . Pictures were taken at room temperature with a CoolSNAP cf2; Photometrics camera . All pictures were exported in preparation for printing using Photoshop ( Adobe ) . MetaMorph software ( MDS Analytical Technologies ) was used to quantify the mean and s . e . m . of integrated intensities per cell as described in [46] . The 2–4 most anterior intestinal cells were used for the quantification of H3K27me3/GFP and more than 20 cells from at minimum of 10 animals for each genotype ( N2+UTX-1::GFP , N2+UTX-1DD::GFP and GFP-negative siblings ) were analyzed in two independent experiments . Only animals showing good H3K27me3 signal in the gonads , as indication of successful immunostaining , were used for quantification . Statistical calculations were performed using the Graphpad Prism software package ( GraphPad Prism version 4 . 00 for Windows , GraphPad Software , San Diego California USA , www . graphpad . com ) . Distribution of data was assessed using three different normality tests: KS normality test , D'agostino & Pearson Normality test and Shapiro-Wilk normality test . When data were normally distributed according to these tests , parametric statistics were applied ( t-test ) , otherwise non-parametrical statistical analysis ( Mann-Whitney U test ) was performed . When comparing more than two groups ANOVA tests were applied . For all tests p≤0 . 05 was considered significant . For immunostaining , animals were fixed and permeabilized as described [47] . Polyclonal anti-H3K27me3 ( Upstate 07-449 ) and polyclonal anti-UTX-1 ( Eurogentec , clone 3917 , this study ) were used . Secondary antibodies were: goat anti-mouse IgG ( Alexafluor 488 ) ; goat anti-rabbit IgG ( Alexafluor 594 ) , both purchased from Invitrogen . DAPI ( Sigma , 2 ug/ul ) was used to counter-stain DNA . Eggs immunofluorescence was performed by freeze crack method , adding eggs to polylysine treated slides . After freezing at −80°C for 30 minutes , the cover slip was removed and embryos were fixed in methanol at −20°C for 10 min . Primary antibody was incubated overnight at 4°C in a humid chamber and secondary antibody was incubated 1 h at room temperature . Washes were in PBS/tween 0 . 2% . Mounting medium for fluorescence with DAPI ( Vectashield H1200 ) was used . Generation of transgenic strain utx-1+UTX-1::GFP has been described earlier in Materials and Methods . Total protein extracts was obtained by grinding a frozen pellet of mixed eggs and adults with a mortar and pestle into powder , the latter was resuspended in IP buffer containing 300 mM KCl , 0 . 1% Igepal , 1 mM EDTA , 1 mM MgCl2 , 10% glycerol , 50 mM Tris HCl ( pH 7 . 4 ) and protease inhibitors . GFP-Trap beads ( Chromotek ) were used to precipitate GFP-tagged proteins from this lysate . Approximately 200 mg of total proteins was used for the pulldown in IP buffer . Following incubation and washes with the same buffer , proteins were eluted with acidic glycine ( 0 . 1 M [pH 2 . 5] ) , resolved on a 4–12% NuPage Novex gel ( Invitrogen ) , and stained with Imperial Protein Stain ( Thermo Scientific ) . The gel was sliced into 21 bands across the entire separation range of the lane . Cut bands were reduced , alkylated with iodoacetamide , and in-gel digested with trypsin ( Promega ) as described previously [48] , prior to LC/MS-MS analysis . Peptide identification was performed on an LTQ-Orbitrap mass spectrometer ( Thermo Fisher Scientific , Germany ) coupled with an EASY-nLC nanoHPLC ( Proxeon , Odense , Denmark ) . Samples were loaded onto a 100 µm ID×17 cm Reprosil-Pur C18-AQ nano-column ( 3 µm; Dr . Maisch GmbH , Germany ) . The HPLC gradient was from 0 to 34% solvent B ( A = 0 . 1% formic acid; B = 95% MeCN , 0 . 1% formic acid ) over 30 minutes and from 34% to 100% solvent B in 7 minutes at a flow-rate of 250 nL/min . Full-scan MS spectra were acquired with a resolution of 60 , 000 in the Orbitrap analyzer . For every full scan , the seven most intense ions were isolated for fragmentation in the LTQ using CID . Raw data were viewed using the Xcalibur v2 . 1 software ( Thermo Scientific ) . Data processing was performed using Proteome Discoverer beta version 1 . 3 . 0 . 265 ( Thermo Scientific ) . For database search we included both Mascot v2 . 3 ( Matrix Science ) and SEQUEST ( Thermo Scientific ) search engines . Database of C . elegans protein sequences was downloaded from Uniprot . Trypsin was selected as digestion enzyme and two missed cleavages were allowed , carbamidomethylation of cysteines was set as fixed modification and oxidation of methionine as variable modification . MS mass tolerance was set to 10 ppm , while MS/MS tolerance was set to 0 . 6 Da . Peptide validation was performed using Percolator and peptide false discovery rate ( FDR ) was set to 0 . 01 . For additional filtering , maximum peptide rank was set to 1 and minimum number of peptides per protein was set to 2 . Protein grouping was performed , in order to avoid presence of different proteins identified by non-unique peptides . We manually investigated whether the protein listed to represent the protein group was the most characterized in terms of sequence coverage and number of peptides identified . For co-immunoprecipitation assays , frozen eggs ( prepared by hypochlorite treatment ) were reduced into powder using a mortar and pestle . The powder was resuspended in IP buffer ( described in GFP pulldown section ) and 5–10 mg was incubated with Protein G agarose beads ( Upstate ) overnight at 4°C . Soluble fraction was collected and incubated with EZview anti-HA affinity gel beads ( Sigma Aldrich ) during 2 h at 4°C . The immunoprecipitates and the protein G beads were washed five times in IP buffer , boiled in SDS-sample buffer and analyzed by SDS-PAGE followed by western blotting . Antibodies used in those experiments were: anti-HA ( Covance HA . 11 , clone 16B12 ) , anti-GFP ( Roche , 11814460001 ) and anti-UTX-1 ( Eurogentec , clone 3917 , this study ) . Quantification of western blots was performed using ImageJ program ( National Institutes of Health , USA ) . Eggs from indicated strains were grinded to powder , resupended in IP buffer ( 50 mM Tris-HCl pH 7 . 4 , 300 mM KCl , 1 mM MgCl2 , 1 mM EDTA , 0 . 1% Igepal and complete protease inhibitors [Roche] ) and incubated on wheel for 30 min at 4°C . Protein extracts were recovered by centrifugation at 20 , 000 g , 30 min at 4°C and clarified by ultracentrifugation at 627 , 000 g for 30 min at 4°C . Fresh extracts were fractionated on a Superose 6 HR 10/300 GL column ( GE Healthcare ) equilibrated in IP buffer . Size exclusion chromatography was performed on a fast protein liquid chromatography ( FPLC ) system and an ÄKTA purifier ( GE Healthcare ) . Elution profiles of blue dextran ( 2 , 000 kDa ) , thyroglobulin ( 660 kDa ) and bovine serum albumin ( 66 kDa ) were used for calibration . Fractions of 1 ml were collected and precipitated with 25% trichloroacetic acid and then centrifuged at 20 , 000 g for 10 min at 4°C . Pellets were washed two times in cold acetone , air dried , and resuspended in loading buffer for Western blot analysis .
|
Chromatin organization influences gene expression , and its regulation is crucial to achieve correct cellular differentiation and development in multicellular organisms . Histone demethylases are among several factors responsible for regulating chromatin dynamics . Here we report on the biological functions of UTX-1 , the C . elegans homologue of the mammalian histone demethylase UTX , which specifically catalyzes the demethylation of di- and tri-methylated lysine 27 of histone H3 ( H3K27me2/3 ) . Indeed , we demonstrate that UTX-1 regulates global levels of H3K27me2/3 in C . elegans , a mark generally associated with silencing of gene expression . We also show that utx-1 is an essential gene that is required for correct embryonic and postembryonic development . Specifically , the loss of utx-1 results in developmental defects , sterility , and embryonic lethality . Surprisingly , our data show that the catalytic activity of UTX-1 is not required for its developmental functions . Our biochemical and genetic analyses indicate that loss of UTX-1 compromises the activity of the SET-16 ( MLL ) complex , which UTX-1 is an integral part of . Taken together , these results demonstrate that UTX-1 plays an essential role in development independent of its enzymatic activity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"animal",
"models",
"biochemistry",
"developmental",
"biology",
"caenorhabditis",
"elegans",
"model",
"organisms",
"organism",
"development",
"protein",
"interactions",
"genetics",
"epigenetics",
"biology",
"proteomics",
"genetics",
"and",
"genomics",
"histone",
"modification"
] |
2012
|
The C. elegans H3K27 Demethylase UTX-1 Is Essential for Normal Development, Independent of Its Enzymatic Activity
|
The Bck2 protein is a potent genetic regulator of cell-cycle-dependent gene expression in budding yeast . To date , most experiments have focused on assessing a potential role for Bck2 in activation of the G1/S-specific transcription factors SBF ( Swi4 , Swi6 ) and MBF ( Mbp1 , Swi6 ) , yet the mechanism of gene activation by Bck2 has remained obscure . We performed a yeast two-hybrid screen using a truncated version of Bck2 and discovered six novel Bck2-binding partners including Mcm1 , an essential protein that binds to and activates M/G1 promoters through Early Cell cycle Box ( ECB ) elements as well as to G2/M promoters . At M/G1 promoters Mcm1 is inhibited by association with two repressors , Yox1 or Yhp1 , and gene activation ensues once repression is relieved by an unknown activating signal . Here , we show that Bck2 interacts physically with Mcm1 to activate genes during G1 phase . We used chromatin immunoprecipitation ( ChIP ) experiments to show that Bck2 localizes to the promoters of M/G1-specific genes , in a manner dependent on functional ECB elements , as well as to the promoters of G1/S and G2/M genes . The Bck2-Mcm1 interaction requires valine 69 on Mcm1 , a residue known to be required for interaction with Yox1 . Overexpression of BCK2 decreases Yox1 localization to the early G1-specific CLN3 promoter and rescues the lethality caused by overexpression of YOX1 . Our data suggest that Yox1 and Bck2 may compete for access to the Mcm1-ECB scaffold to ensure appropriate activation of the initial suite of genes required for cell cycle commitment .
The temporal control of transcription is likely a universal feature of cell cycles , with clear transcriptional programs in yeast , bacteria and metazoans [1]–[4] . Bursts of gene expression in eukaryotes tend to be associated with major cell cycle transitions that are governed by cyclin-dependent kinases ( Cdks ) in association with regulatory subunits called cyclins [5] . Up-regulation of G1-specific forms of the Cdc28 Cdk is rate-limiting for cell cycle commitment . Control over Cdc28 activity is exerted at several levels , including transcriptional induction of cyclin gene expression; the G1/S phase transition features activation of a massive transcriptional program of ∼120 genes , including the G1 cyclin genes CLN1 and CLN2 [6]–[10] . Two heterodimeric transcription factors largely drive the G1/S cluster: SBF ( Swi4 , 6-dependent cell cycle box binding factor , a heterodimer of Swi4 and Swi6 ) binds the so-called SCB promoter element ( Swi4 , 6-dependent cell cycle box ) found upstream of the cyclin genes and cell wall biosynthetic genes , while MBF ( MluI cell cycle box binding factor , a heterodimer of Mbp1 and Swi6 ) preferentially acts through a distinct element called the MCB ( MluI cell cycle box ) found mostly upstream of DNA replication and repair genes . Although the role of SBF and MBF at the G1/S transition is well established , deletion of Swi6 , the common subunit of both complexes , does not cause cell cycle arrest , suggesting alternative pathways for activating G1 transcription . One of these pathways is defined by Bck2 , a poorly understood cell cycle regulator whose deletion causes dramatic cell cycle phenotypes in certain genetic contexts . For example , deletion of both CLN3 , which encodes a cyclin that activates Cdc28 at M/G1 , and BCK2 causes synthetic lethality [11]–[13] . Furthermore , BCK2 is one of only two genes whose overexpression is known to bypass the lethality caused by mutation of all three G1-specific Cdc28 cyclins ( CLN1 , CLN2 and CLN3 ) [11] . Bck2 was discovered almost 20 years ago in screens for high copy suppressors of protein kinase C pathway mutants ( including pkc1 itself ) and G1 cyclin deficiencies [11] , [14] and is thought to have an activating role in transcription of G1/S genes [12] , [13] . Activation of G1/S genes by Bck2 depends partly , but not wholly , on SBF and MBF and Bck2 appears capable of activating G1/S transcription in the absence of Cdc28 , indicating SBF- and MBF-independence [13] . Furthermore , a region in the promoter of CLN2 , termed UAS2 , which completely lacks SCB or MCB elements , is activated by BCK2 overexpression [12] and operates in a CDC28-independent manner [15] . Bck2 also has a role in regulating heat shock genes and adaptation of the protein kinase C pathway MAP kinase Slt2 [16] , although it is not clear whether these roles depend on SBF . Collectively , these findings suggest that Bck2 may operate through an unidentified DNA-binding factor whose activity is Cdc28-independent . The MADS box transcription factor Mcm1 has an important regulatory function at two points in the cell cycle – M/G1 and G2/M . During M/G1 , Mcm1 functions as a critical constituent of a complex that forms on ECB ( Early Cell cycle Box ) elements in promoters of genes expressed at the M/G1 phase transition such as CLN3 and SWI4 [17] . Two related homeodomain proteins , Yox1 and Yhp1 , act as repressors of Mcm1 on ECB elements by physically interacting with Mcm1 and with DNA binding sites next to the Mcm1 site in the ECB element [18] . Mcm1 is also a critical constituent of complexes that form during the G2/M phase transition to control the CLB2 cluster of genes , including the G2 cyclin gene CLB2 and CDC20 , which encodes an activator of the anaphase promoting complex [19] . The CLB2 gene cluster is activated by the Clb-Cdc28 and Cdc5 kinases [20]–[23] and negatively regulated by protein kinase C , Pkc1 [24]; these kinases regulate a promoter-bound complex of Mcm1 , Fkh2 and Ndd1 [25]–[29] . Some CLB2 cluster genes , but not CLB2 itself , contain hybrid elements composed of an Mcm1-binding site flanked by Yox1- and Fkh2-binding sites [18] . In CLB2 cluster genes that contain such elements , Yox1 and Fkh2 compete for binding to DNA-bound Mcm1 despite the spatial separation of their DNA recognition elements [30] . Interestingly , no such juxtaposed binding motifs are obvious in the vicinity of ECB elements and no other binding partners for Mcm1 that positively regulate these genes have been identified . Mcm1 also has roles at the promoters of some other genes , including the G1/S gene CLN2 , where it contributes to the presence of a nucleosome depleted region that is needed for reliable “on/off” expression once per cell cycle [31] . In this study , we show that Bck2 activates expression of the M/G1 genes CLN3 and SWI4 , through an interaction with Mcm1 on ECB elements in the promoters of these genes . Moreover , increased Bck2 dosage leads to decreased Yox1 binding at M/G1 promoters and overproduction of BCK2 rescues the lethality caused by overexpression of YOX1 , indicating that Bck2 and Yox1 may compete for access to Mcm1 on promoters . Consistent with this hypothesis , mutation of a key residue on Mcm1 known to prevent interaction with Yox1 also prevents interaction with Bck2 . In addition , we show that Bck2 interacts with the promoters of the G1/S gene CLN2 and the G2/M gene CLB2 . Our experiments reveal a previously unappreciated function for Bck2 as a cofactor for Mcm1 and suggest a more general role for Bck2 in regulating cell-cycle gene expression .
We reasoned that an exploration of Bck2-interacting proteins might illuminate the function of Bck2 in G1 progression . Bck2 is difficult to work with biochemically ( for example endogenous Bck2 is undetectable by Western blot [16] , [32 , and data not shown] , so we turned to the yeast two-hybrid ( Y2H ) system for our protein interaction screens . Bck2 has a potent transcriptional activation domain , and activates reporter gene expression when fused to the Gal4 DNA binding domain ( DBD ) in a Y2H reporter strain [33] ( Figure 1 ) , a property that has precluded identification of Bck2 binding partners using the Y2H screening method . To discover Bck2 derivatives that might be useful for two-hybrid screening , we created 20 truncations of the BCK2 gene and fused them to the GAL4 DBD ( Figure 1 ) . We first assessed the ability of each truncation construct to activate transcription of two reporter genes , in which the GAL4 UAS is upstream of either lacZ ( Figure S1 ) or ADE2 ( Figure S1B ) . Bck2 residues 662 to 851 were required for transcriptional activation ( fragments 11–20 ) , while the Bck2 N-terminal region had some apparent inhibitory activity ( Figure S1A ) , as suggested by our finding that deletions of the N-terminus resulted in significant increases in reporter gene expression . A construct containing fragment 5 , lacking the first 529 amino acids of Bck2 , was the most potent Y2H auto-activator in the lacZ reporter assay ( Figure S1A ) . Proteins that autoactivate in the Y2H system often have a role in transcription and this property may reflect the biological activity of the protein [33] . To explore the relationship between Y2H autoactivation and biological function , we assessed the ability of each Bck2 fragment to complement the synthetic lethal phenotype of a cln3Δbck2Δ strain ( Figure S1C ) . We discovered that a fragment of Bck2 containing residues 250 to 766 was able to robustly complement the lethality of the cln3Δbck2Δ strain ( Figure 1 , Figure S1C ) . Consistent with the observation that the first 178 residues of Bck2 are not necessary for suppression of the pkc1 lysis phenotype by high-copy BCK2 [14] , a derivative of Bck2 lacking the N-terminal 178 residues was also able to complement the inviability of the cln3Δbck2Δ strain ( Figure S1C ) . Bck2 residues 529 to 851 alone failed to complement the cln3Δbck2Δ phenotype but were sufficient for Y2H auto-activation . This region is also insufficient to suppress the lysis phenotype of pkc1 mutants [14] . We conclude that the central region ( 250 to 766 ) of Bck2 lacking the N- and C-terminal ends is sufficient to complement essential in vivo functions of Bck2 and that this essential function is separable from the Y2H auto-activation activity of the Bck2 protein . Thus far , no known protein interaction partners of Bck2 easily explain the cell cycle transcription phenotypes associated with deletion of BCK2 . To carry out a Y2H screen , we decided to use the largest Bck2 construct that did not auto-activate the ADE2 Y2H reporter gene , but complemented the inviability of the cln3Δbck2Δ strain ( fragment 11; Figure 1 ) . We chose this fragment for our screen since complementing regions often contain important protein-protein interaction domains . For example , the minimal region of the Ada2 protein required for complementation is the same region required for physical interaction with Gcn5 and Ada3 [34] . We used the ORFeome Y2H screening method [35] to discover potential Bck2-interacting proteins . We identified six proteins that interacted with Bck2: Mcm1 , Yap6 , Tpd3 , Std1 , Mth1 , and Mot3 ( Figure 2 and data not shown ) . With the exception of Tpd3 , all of these proteins are transcriptional regulators that act proximal to DNA . As noted in the introduction , Mcm1 is a member of a class of MADS box transcription factors found in all eukaryotic organisms [36]–[38] and we explore the Mcm1-Bck2 interaction in detail below . Four of the other Bck2 partners we identified in our screen were also DNA binding proteins that had roles in ion homeostasis and nutrient sensing: ( 1 ) Yap6 is a basic leucine zipper ( bZIP ) transcription factor that activates a number of genes involved in sodium and lithium tolerance [39] , [40]; ( 2 ) Std1 and Mth1 are both controllers of glucose-regulated gene expression [41] and are required for transcriptional repression of HXT ( hexose transport ) genes [42] , [43]; ( 3 ) Mot3 is a Zn-finger transcription factor that activates a number of cell wall genes [44] , and represses transcription of the DAN/TIR group of genes that encode cell wall mannoproteins during anaerobic growth [44] . The only protein that is not a transcription factor , Tpd3 , is the scaffold subunit A of the heterotrimeric protein phosphatase 2A ( PP2A ) [45] , which has roles at several points in the cell cycle and is involved in the TOR pathway for nutrient sensing [46] . We did not identify SWI4 or MBP1 in our Y2H screen , but a direct test revealed a weak interaction between Bck2 and Swi4 ( Figure 2 ) . The identification of genes with known roles in transcription and nutrient response is consistent with the apparent functions of Bck2 , and validates our two-hybrid screen as a tool for identifying Bck2-interacting proteins . Given the clear roles for Mcm1 in cell-cycle-dependent gene expression , we chose to focus our follow-up analysis on the Mcm1-Bck2 interaction . In addition to the Y2H interaction between Bck2 and Mcm1 , several observations from earlier studies implicate Bck2 in the activation of Mcm1 target genes in M/G1 phase: ( 1 ) mRNA from the M/G1 gene SWI4 accumulates much more slowly in bck2Δ than WT cells in synchronized cultures , whereas SWI4 is upregulated in cells that overexpress BCK2 [12]; ( 2 ) high-copy BCK2 stimulates the expression of the Mcm1-dependent reporter gene , P-lacZ [47]; ( 3 ) overexpression of BCK2 causes increased transcription of CLN3 and SWI4 by microarray analysis [48] . However , up to now , no direct connection between BCK2 and M/G1 genes has been established . To evaluate the significance of the Mcm1-Bck2 physical interaction in vivo , we first assessed the effect of BCK2 deletion on expression of lacZ reporter genes whose expression was dependent on either multiple Mcm1-binding sites ( 4 x P-sites ) [47] or the upstream activating sequences of four Mcm1-regulated genes expressed in M/G1 phase – CLN3 , CDC6 , CDC47 , SWI4 [49] ( Figure 3A ) . In these plasmid reporter assays , deletion of BCK2 had no effect on expression of a control ACT1-lacZ reporter gene . However , we saw a pronounced reduction in expression of the CLN3-lacZ , CDC6-lacZ , CDC47-lacZ , SWI4-lacZ , and P-lacZ reporter genes in the bck2Δ strain ( Figure 3B ) . These results suggest a role for BCK2 in M/G1 gene expression . To verify the results of the reporter gene assays , we next examined endogenous levels of Mcm1 target gene expression in a bck2Δ strain ( Figure 4 ) . Since Mcm1 controls CLN3 and SWI4 transcript accumulation at a very early point in G1 , we synchronized cultures in mitosis with a cdc20-3 temperature-sensitive allele and released them into the subsequent cell cycle . Cells in this experiment were slow growing and enriched in large-budded cells that precluded FACS analysis of cell cycle synchrony ( data not shown ) . However , CLN2 transcription was highly periodic in both WT and bck2Δ cells , indicating that these cultures were synchronously released from the mitotic block . Consistent with previous reports , CLN2 transcript was reduced in the bck2Δ strain [11] , [12] while levels of a control transcript ( ALG9 ) were unaffected . Strikingly , the accumulation of CLN3 and SWI4 mRNAs was significantly reduced in bck2Δ cells , and peak expression was also delayed , at least for the CLN3 transcript . In wild-type cells , CLB2 mRNA peaked after G1 transcripts as expected . However , in bck2Δ cells CLB2 transcripts were delayed and only began to accumulate near the end of the time-course , after the peak of CLB2 expression seen in wild-type cells . Thus , BCK2 is required for the appropriate expression of the M/G1 genes CLN3 and SWI4 , the G1/S gene CLN2 , and the G2/M gene CLB2 . The observation that Bck2 is required for proper transcription of M/G1 genes that contain Mcm1-binding sites suggested that Bck2 may function through ECB elements . To test this hypothesis , we assayed the effect of BCK2 deletion on expression of lacZ reporter plasmids containing either a WT CLN3 promoter that has intact ECB elements or a mutated CLN3 promoter that lacks functional ECB elements [50] . Deletion of BCK2 in combination with mutation of ECB elements caused a level of reporter gene expression similar to that seen with either perturbation alone ( Figure 5A ) , indicating that Bck2 acts through ECB elements in order to control expression of M/G1 genes . To gather more evidence that Bck2 works through ECB elements , we next tested the effects of BCK2 overexpression on ECB-containing reporter gene expression . For this experiment , we used a lacZ reporter gene driven by a version of the CLN3 promoter in which ECB elements were mutated . As previously seen [50] , when ECB elements were mutated , expression of CLN3pr-lacZ in wild-type cells was substantially reduced ( Figure 5B ) . Overexpression of BCK2 induced expression of the wild-type CLN3pr-lacZ reporter gene approximately 3-fold , in a manner that was largely dependent on intact ECB elements ( Figure 5B ) . We conclude that ECB elements are required for Bck2 to maximally activate transcription through the CLN3 promoter . To substantiate the requirement of ECB elements in transcriptional activation of M/G1-regulated genes by overproduced Bck2 , we next assayed the effects of BCK2 overexpression on expression of the endogenous CLN3 and SWI4 genes . We compared the expression of CLN3 and SWI4 in a wild-type strain to a strain where ECB elements in the promoters of both CLN3 and SWI4 were mutated ( cln3 ( ecb ) swi4 ( ecb ) ) [49] . Consistent with our reporter gene assays , overproduction of Bck2 increased CLN3 and SWI4 transcript levels in a WT strain , but not the cln3 ( ecb ) swi4 ( ecb ) mutant strain ( Figure 5C ) , indicating that Bck2 functions through ECB elements in endogenous M/G1 promoters . As previously seen , overproduction of Bck2 also increased CLN2 expression in WT cells [12] , [51]; however , this induction was entirely independent of the ECB elements in the CLN3 and SWI4 gene promoters ( Figure 5C ) . This result suggests that the induction of CLN2 transcription by overexpressed BCK2 is not an indirect consequence of increased CLN3 and SWI4 expression . Similarly , the induction of CLB2 expression by overexpressed BCK2 [12] , [51] was also independent of the ECB elements in the CLN3 and SWI4 promoters ( Figure 5C ) , again suggesting that the CLB2 induction is not an indirect effect of defects in M/G1 phase gene expression . Finally , overexpressed BCK2 did not alter expression of ALG9 , a non-ECB containing gene , indicating that Bck2 is not an activator of global transcription . Together , our analyses of the effects of BCK2 deletion and overexpression show that Bck2 activates CLN3 and SWI4 transcription in an ECB-dependent manner , while also promoting expression of G1/S ( CLN2 ) and G2/M-regulated ( CLB2 ) genes by a mechanism that does not depend on its effect on early G1 genes . Since Bck2 functions through ECB elements ( Figure 5 ) and physically interacts with Mcm1 ( Figure 2 ) , we next asked if Bck2 localized to the promoter regions of M/G1-phase genes . We first used chromatin immunoprecipitation ( ChIP ) with a strain carrying a TAP-tagged allele of Bck2 to assess association of Bck2 with various promoters . We detected a reproducible enrichment of promoter DNA in the Bck2 ChIP , but the signal was very low relative to Swi4 or Mcm1 ChIPs ( data not shown ) . To improve our assay , we repeated the ChIP experiment using a strain in which a FLAG-tagged derivative of Bck2 was conditionally overproduced ( Figure 6A ) . Under inducing conditions ( galactose ) , Bck2-FLAG IPs were enriched in CLN2 , CLN3 and SWI4 promoter DNA relative to non-inducing conditions ( raffinose ) ( Figure 6A ) or vector control ( data not shown ) . The enhanced enrichment of CLN3 promoter DNA compared to SWI4 promoter DNA in Bck2 IPs likely reflects the presence of more ECB elements in the CLN3 promoter ( 6 versus 1 ) . Association of Bck2 with the CLN3 and SWI4 promoters was entirely dependent on the presence of ECB elements , while association with the CLN2 promoter was unaffected , consistent with our gene expression analysis ( Figure 5 ) . Our findings are supported by a recent study that identified Bck2 as a constituent of DNA-bound complexes containing Mcm1 [52] . We conclude that Bck2 localizes to the promoters of CLN3 and SWI4 in a manner that depends on ECB elements . Bck2 also localized to the CLN2 promoter in our ChIP assays . The CLN2 promoter contains three SCB elements , which are required for cell-cycle-specific gene expression , and two Mcm1 binding sites , which contribute to the presence of a nucleosome depleted region but have no effect on average expression levels or cell-cycle-specific expression [31] . We assayed the ability of Bck2 to localize to the CLN2 promoter in strains that contained mutations in either set of elements [31] . Mutation of either the SCBs ( scb* ) or the Mcm1-binding sites ( mcm1* ) led to a modest reduction in Bck2 localization to the CLN2 promoter ( Figure 6B ) , suggesting that Bck2 may be recruited to CLN2 by either SBF or Mcm1 . The SBF-dependence of Bck2 recruitment to the CLN2 promoter was further supported by a reduced Bck2 ChIP in the absence of SWI4 or SWI6 ( Figure S2 ) . Since we saw that BCK2 was required for proper expression of CLB2 ( Figure 4 ) , and that overproduced Bck2 led to increased CLB2 RNA levels ( Figure 5C ) , we next asked whether Bck2 could also localize to the CLB2 promoter . Mcm1 binds to the CLB2 promoter in cooperation with another DNA-binding protein Fkh2 , which is activated by Ndd1 [25]–[29] . Our ChIP experiments revealed that Bck2 localized to the CLB2 promoter with an efficiency comparable to that seen with the CLN3 promoter ( Figure 6C ) , and the binding was partially dependent on FKH2 ( Figure S3 ) . As noted earlier , the closely related homeodomain proteins Yox1 and Yhp1 act as repressors of Mcm1 by interacting directly with Mcm1 at DNA binding sites adjacent to the actual Mcm1 DNA binding site within the ECB [18] . Activation of Mcm1 on ECB elements correlates with removal of Yox1 from ECB elements , while deletion of YHP1 has little effect on the level or periodicity of gene expression , and Yhp1 is not part of the predominant complex at ECB elements [18] . These observations suggested that Bck2 may activate CLN3 and SWI4 transcription through ECB elements by promoting the removal of Yox1 . To test this hypothesis , we first assessed Yox1 binding to ECB elements within the CLN3 promoter when Bck2 levels were elevated . Overexpression of BCK2 significantly reduced the amount of Yox1 associated with the CLN3 promoter ( Figure 7A; left panel ) , implicating Bck2 in Yox1 removal . Yhp1 was not localized to the CLN3 promoter to the same extent as Yox1 , nor was the association affected by BCK2 overproduction , consistent with a secondary role for Yhp1 [18] . Consistent with previous reports that Mcm1 remains localized to promoters throughout the cell cycle [18] , we observed that Mcm1 localization was not significantly affected by BCK2 dosage ( Figure 7A , right panel ) . These experiments suggest that Bck2 overproduction may lead to the displacement of Yox1 repressor from the CLN3 promoter , which is correlated with activation of M/G1-regulated genes [18] . Overexpression of YOX1 is toxic , likely because high levels of Yox1 cause constitutive repression of its target genes [18] . To test the model that Bck2 may compete with Yox1 binding to Mcm1 at ECB elements , we asked if the toxicity caused by YOX1 overexpression was suppressed by concurrent overexpression of BCK2 . Previous work on the CLB2 gene cluster , which is expressed at the G2/M phase transition , has shown that Mcm1 acts as a common scaffold for recruitment of Yox1 and the forkhead protein Fkh2 [30] . These physical interactions with Mcm1 are mutually exclusive and are mediated by distinct Yox1 and Fkh2 DNA binding elements that flank a central Mcm1 DNA binding site . When bound to Mcm1 on promoters , Fkh2 recruits the positively acting co-regulator Ndd1 in order to activate the CLB2 cluster genes [26] . Constitutive overexpression of YOX1 inhibits Ndd1 binding to the Yox1-regulated SPO12 promoter [30] , consistent with a mechanism based on mutual exclusivity between an activator and repressor . Our observation that overproduction of Bck2 leads to reduced Yox1 on the CLN3 promoter ( Figure 7A ) suggested that the YOX1 dose-lethality phenotype might be suppressed by concurrently overproducing BCK2 . Indeed , we observed that overexpression of BCK2 was able to significantly suppress the lethality caused by overexpressing YOX1 ( Figure 7B ) . Overexpression of CLN3 failed to rescue the YOX1 overexpression phenotype , suggesting that the BCK2 rescue does not simply reflect an indirect effect of reduced G1 phase . Although the suppression of YOX1 toxicity by overexpressed BCK2 does not show that the effects are direct , it is consistent with a competitive relationship between Bck2 and Yox1 for interaction with the Mcm1 scaffold . Mcm1 contains a hydrophobic pocket found on the surface of the MADS DNA-binding domain , and mutation of this pocket by the introduction of a V69E mutation disrupts interaction with Fkh2 [53]–[55] , which prevents binding of both Yox1 and Fkh2 to Mcm1 [30] . We hypothesized that the competitive binding mechanism that allows Mcm1 to activate genes transcribed at the G2/M transition may also function for genes transcribed at the M/G1 transition . Specifically , we wondered whether Bck2 might activate Mcm1 at ECB elements through competition with Yox1 for binding to Mcm1 . Consistent with our hypothesis , we found that the Y2H-based interaction between Bck2 and Mcm1 was abolished in the Mcm1V69E mutant ( Figure 7C ) . We conclude that Bck2 may act to remove Yox1 by a competitive binding mechanism similar to that seen at G2/M phase promoters [30] .
Our work identifies a previously unappreciated role for Bck2 in the regulation of M/G1-specific transcription in budding yeast , and also reveals functions for Bck2 in regulating late G1 and G2/M genes ( Figure 8 ) . In contrast to the CLB2 gene cluster , no positively acting partner protein had yet been found that cooperates with Mcm1 to regulate M/G1-expressed genes . We describe several observations showing that Bck2 functions to control Mcm1 activity on promoter elements to ensure the proper regulation of early G1 events . First , Bck2 is required to activate expression of the M/G1 genes SWI4 , CLN3 , CDC6 , and CDC47 . Second , we show a requirement for intact Mcm1-binding ECB elements within the promoters of CLN3 and SWI4 , consistent with the physical interaction between Bck2 and Mcm1 that we see by Y2H . Third , Bck2 localizes to early G1 promoters in an ECB-dependent manner . Our finding that Bck2 is needed for expression of Mcm1-regulated genes in M/G1 phase may explain previous observations that Bck2 activates G1/S gene expression in a pathway distinct from that involving Cln-Cdc28 activation of promoter-bound SBF/MBF . For example , high-copy BCK2 activates SBF/MBF target genes in a cdc28-4 mutant [13] and suppresses the G1-arrest of a cln1Δcln2Δcln3Δ strain [11] . These activities likely reflect increased SWI4 expression , because Bck2 activates SWI4 transcription [12] , [48] , and high-copy suppression by BCK2 of the G1-arrest of a cln1Δcln2Δcln3Δ strain requires SWI4 [11] . Importantly , SWI4 is induced in a cdc28-13 mutant [7] . There is no evidence that Cdk activity is required for activation of M/G1-expressed genes . Bck2 has also been shown to activate G1/S genes , in a manner that is both partially dependent and partially independent of SBF/MBF activity . For example , overproduction of Bck2 stimulates expression of an SCB-lacZ reporter gene , activation of which is strictly dependent on SBF [12] . Large-scale mass spectrometry experiments reveal an interaction between Bck2 and Swi4 [56] and we see a weak Y2H interaction between BCK2 and SWI4 ( Figure 2 ) . In our experiments , localization of Bck2 to the CLN2 promoter was modestly reduced when the SCB sites were mutated ( Figure 6B ) and Bck2 localization was partially dependent on SWI4 and SWI6 ( Figure S2 ) , suggesting that SBF has a role in recruiting Bck2 . However , consistent with an SBF-independent activity , Bck2 can activate several natural SBF/MBF target gene promoters in the absence of either SWI4 or MBP1 [13] , or the elements SBF/MBF bind [12] . In our experiments , mutation of the Mcm1-binding sites in the CLN2 promoter also modestly reduced Bck2 binding ( Figure 6B ) , suggesting that Mcm1 also contributes to Bck2 recruitment to the CLN2 promoter . Since SCB elements are much more important for proper CLN2 expression than Mcm1-binding sites [31] , it is likely that SBF has a more important role than Mcm1 in recruiting Bck2 to the CLN2 promoter . We present several lines of evidence that suggest a role for Bck2 in activation of G2/M genes . First , a bck2Δ strain has delayed CLB2 expression ( Figure 4 ) . Second , overexpression of BCK2 leads to increased expression of CLB2 [12] ( Figure 5C ) , and other G2/M genes [51] . Third , we show that Bck2 localizes to the CLB2 promoter ( Figure 6C ) , in a manner dependent on FKH2 ( Figure S3 ) . Activation of G2/M genes is controlled by Ndd1 , which is recruited to promoters by binding Fkh2 [25]–[29]; it is not clear why FKH2 is involved in Bck2 localization to the CLB2 promoter . One possibility is that Bck2 binds to both Fkh2 and Mcm1 at G2 promoters , similar to its binding to SBF and Mcm1 at G1/S promoters; however , we did not detect an interaction between BCK2 and FKH2 in a pairwise Y2H assay ( data not shown ) . Regardless , the mechanism of action of Bck2 at G2/M promoters must differ from that at M/G1 promoters where Bck2 appears to act on Mcm1 alone . Unlike the G2/M promoters , M/G1 promoters [6] do not contain DNA binding sites for Fkh2 [29] or other positive regulators , indicating that induction of M/G1 [18] genes depends on a positively-acting protein that presumably acts through Mcm1 . Mcm1 , together with class-specific regulators , controls the expression of several different groups of genes in addition to the cell-cycle-regulated genes discussed above: ( 1 ) activation of genes involved in mating together with Ste12; ( 2 ) regulation of cell-type-specific genes together with α1 or α2; and ( 3 ) control of genes involved in arginine metabolism together with Arg80 ( for review see [57] ) . Other observations suggest that Bck2 may be involved in regulating expression of other classes of Mcm1-dependent genes . First , BCK2 was identified as a gene whose overexpression strongly induced the expression of FUS1 reporter genes [58] . Second , a more recent study examining transcriptional profiles found that overexpression of BCK2 led to expression of cell-cycle-regulated genes from multiple cell-cycle stages , consistent with our findings [51] . Furthermore , overexpression of BCK2 induced a large number of genes involved in mating , but not genes involved in cell type or nitrogen metabolism [51] . Indeed these authors suggested that Bck2 may elicit gene expression via Ste12-Mcm1 . We have shown that Bck2 acts through Mcm1 to promote expression of three classes of cell cycle genes; we suggest that Bck2 likely regulates mating genes in a similar manner . We have shown in this work that Bck2 regulates cell cycle gene expression . We suggest that the role of Bck2 may be to fine-tune expression of different classes of Mcm1-regulated genes . How might Bck2 itself be regulated ? Bck2 is rich in serine and threonine residues and has been shown to be phosphorylated at multiple sites at different stages of the cell cycle [59] , [60] . Bck2 has been linked to several cell-cycle-regulating kinases and phosphatases in genetic studies ( the Protein Kinase C pathway [16] , [48] , Sit4 [12] , [61] , Cdc28 [13] , [48] , Cbk1 [16] ) and in mass spectrometry and phosphopeptide analyses ( Cdc15 [62] , Cdc28 [60] , Fus3 and Kss1 [56] ) . Thus Bck2 may play a role in linking detection of nutrients or other conditions that affect cell cycle progression to Mcm1-dependent gene expression ( Figure 8 ) . For example , during early G1 phase , yeast cells assess nutrient status , in part through the Tor pathway and the phosphatase Sit4 [63] . BCK2 has been proposed to function in the SIT4 pathway of CLN activation [12] , [61] . Both Tor and Sit4 signaling are also required for proper M/G1 gene expression [63] , [64] , and like tor1Δ cells and sit4Δ cells , bck2Δ cells are rapamycin sensitive [16] , [63] , [65] . Thus Bck2 may be part of the signal transduction pathway linking nutrient status to passage through M/G1 of the cell cycle . Consistent with a potential role for Bck2 in integrating environmental and cell cycle signals , our Y2H screen identified several transcription factors and regulators with established functions in regulating gene expression in response to various nutrients or stresses , including nutrient sensing through the TOR pathway ( see Results and Figure 2 ) . How Bck2 may work together with various proteins to regulate nutrient and stress responses remains to be determined . Based on the high level of similarity between budding yeast Mcm1 and human Serum Response Factor ( SRF ) [30] , [54] , [55] , [66]–[68] , the Bck2 protein might represent a budding yeast analog of a specific SRF co-activator in mammalian cells . For example , the myocardin/MKL family members are SRF co-activators which are enriched at target promoters in serum-rich medium [69] . Our discovery of Bck2 as a cofactor for Mcm1 , coupled with accumulating evidence for a role for Bck2 in nutrient sensing , suggests that Bck2 binding may also increase under nutrient-rich conditions . Further studies will be required to firmly test the analogy between the Mcm1-Bck2 and SRF-MKL pathways and to illuminate how Bck2 may link gene expression , the cell cycle machinery and environmental signals to ensure appropriate cell proliferation .
Yeast strains used in this study ( Table 1 ) were derivatives of either S288C or W303 , with the exception of the Y2H strains . Plasmids are described in Table 2 . Yeast cultures were grown in YEP ( 1% yeast extract , 2% bactopeptone ) supplemented with 2% glucose . Synthetic minimal medium supplemented with the appropriate nutrients was used to select for plasmid maintenance and gene replacements . Yeast transformation and general manipulation of yeast cells were performed using standard techniques . Fragments of the BCK2 gene were amplified from genomic DNA using PCR primers designed to be compatible with the Gateway system of recombinational cloning ( Table S1 ) . All PCR products were recombined into the donor vector pDONR201 using the BP clonase II system ( Invitrogen ) and positive clones were fully sequence-confirmed . BCK2 truncations within pDONR201 were recombined into the destination vector pDEST-DB using the LR clonase II system . Quantitative β-galactosidase assays were performed as follows . Exponentially growing cells at an optical density at 600 nm of 0 . 2 to 0 . 25 were harvested . Extracts were prepared by vortexing the cells in 1 ml Z-buffer ( 0 . 1 M NaPO4 [pH 7 . 0] , 0 . 01 M KCl , 1 mM MgSO4 , 4 mM 2-mercaptoethanol ) +20 µl 0 . 1% SDS+40 µl chloroform for 45 seconds . After 5 min , 0 . 2 ml of o-nitrophenylgalactoside ( Sigma; at 4 mg/ml in Z buffer ) solution was added and the reaction was incubated at 30°C until a slight yellowing was observed . The reactions were stopped by the addition of 0 . 5 ml of 1 M Na2CO3 , and the samples were centrifuged for 3 minutes at 13 , 000× g . The A420 of the supernatant was determined . β-Galactosidase units were calculated using the following formula: units = ( A420 ) * ( 1000 ) / ( time of reaction , in minutes ) * ( volume of extract in assay = 1 ml ) * ( cellular concentration in OD600 values ) . For each time point , the assays were performed on three separate cultures , and the average is reported . Qualitative β-galactosidase overlay assays were performed as described [70] . Transformants of a cln3Δbck2Δ pGAL-CLN3 -URA3 strain ( BY3015 ) bearing ADH1-GAL4 DBD-BCK2-LEU2 ( 1–20 ) plasmids were grown in plasmid selective medium , prepared at equivalent optical density and spotted in serial 10-fold dilutions onto plasmid selective medium containing either galactose lacking 5′FOA or glucose+5′FOA , and incubated for 48 h at 30°C . The Y2H ORFeome method [71] was used to screen for Bck2-interacting proteins by mating a prey strain with a bait strain . The AD ORFeome ( prey strain ) is a pooled collection of AD-ORF plasmids [72] in a 96-well format . The pooled AD ORFeome [72] was grown in a 96-well culture block containing 600 µl per well of SD – Trp media at 30°C for 48 hours . Five µl of AD ORFeome culture were spotted from the 96-well culture block onto large YPD plates and allowed to dry . Next , 5 µl of a DBD ORF strain culture ( bait strain transformed with a single DBD-ORF plasmid ) was dispensed directly onto the ORFeome strain spots . Five diploid control strains were spotted onto the same plate at an empty location in order to ensure the quality of selection plates and to help evaluate the phenotype of interactions . The plate was incubated at 30°C until growth of spots was apparent before replica plating onto SD – Leu – Trp and grown for 3 days at 30°C in order to select only diploid yeast . This plate was replica plated onto a final selection plate containing SD – Leu – Trp – Ade and grown for 5 days until foci were observed . At least 3 foci per spot were picked , and subjected to colony PCR using primers homologous to the sequences flanking the ORF within the AD plasmid . Yeast cells were scraped from plates into 30 µl of lysis solution ( 2 . 5 mg zymolyase in 1 ml 1 M sorbitol ) , and incubated at 37°C for 15 minutes then 95°C for 5 minutes before addition of 120 µl of ddH2O . PCR reactions were performed in 25 µl volumes containing 5 µl of the yeast cell preparation described . PCR reactions were performed using 5 minute extension times in order to ensure that large ORF inserts were isolated . PCR reactions were electrophoresed on agarose gels , purified using the PureLink kit ( Invitrogen ) and sent for sequencing analysis . Yeast transformants carrying ADH1-GAL4 DBD ( vector; LEU2 ) or ADH1-GAL4 DBD-BCK2 Fragment 11 ( Bck2 ) in a two-hybrid bait strain ( Y8930 ) were mated to yeast transformants of a two-hybrid prey strain ( Y8800 ) bearing specific ADH1-GAL4-AD-ORF-TRP plasmids . Diploids were selected by streaking cells onto double plasmid selection medium ( SD – Leu – Trp ) . Diploids of equivalent optical density were spotted in serial 10-fold dilutions on double plasmid selection medium ( SD – Leu – Trp ) , or medium where growth was proportional to transcription of the ADE2 gene ( SD – Leu – Trp - Ade ) , and incubated for 48 h at 30°C . Six diploid strains carrying different combinations of AD and DBD ORF-fusions [72] , [73] were used as a spectrum of positive and negative controls . A cdc20-3 temperature-sensitive strain ( BY4896 , Table 1 ) was grown in YPD medium at 21°C , arrested in M phase by incubation at 37°C for 3 . 5 hours , and released into the cell cycle by transferring the culture back to 21°C . Arrest was determined by visualization of large budded cells under a light microscope . Total RNA was isolated by phenol-chloroform extraction and further purified using the RNeasy kit ( Qiagen ) . RNA was transcribed into cDNA using the Superscript II Reverse Transcriptase kit ( Invitrogen ) and RNA was then removed by addition of NaOH . Reactions were run on the ABI 7500 system ( Applied Biosystems ) using standard Q-PCR conditions . Data were analyzed using ABI7500 system software . VIC and FAM labelled fluorogenic primers ( ABI ) , used to detect CLN2 , ALG9 , CLN3 , SWI4 , BCK2 , CLB2 and ACT1 cDNA , are described in Table S1 . To construct the Mcm1V69E yeast two-hybrid prey plasmid , the Mcm1WT prey plasmid was subject to in vitro mutagenesis using the QuikChange site-directed mutagenesis kit ( Stratagene ) . Yeast strains transformed with pGAL-FLAG plasmids were grown in raffinose-containing minimal media overnight , and then grown separately in raffinose- ( non-inducing conditions ) or galactose-containing medium ( inducing conditions ) to mid-log phase . Cultures were harvested and anti-FLAG ChIPs were analyzed for CLN2 , CLN3 , CLB2 and SWI4 promoter DNA by TaqMan Q-PCR ( Applied Biosystems ) using primers with homology to target gene promoter DNA ( Table S1 ) . Enrichment of promoter DNA was determined relative to non-promoter DNA from an untranscribed region of chromosome II ( Table S1 ) .
|
Cell-cycle-dependent gene expression is a universal feature of cell cycles , with clear transcriptional programs in yeast , bacteria , and metazoans . At the M/G1 transition , many of the up-regulated genes encode key regulators of DNA replication ( CDC6 ) and cyclins that initiate the events of cell cycle commitment ( PCL9 , CLN3 ) . The promoters of genes activated at M/G1 contain a cis-regulatory sequence called the early cell cycle box ( ECB ) , which is bound by the MADS-box transcription factor Mcm1 , as well as the repressor Yox1 or Yhp1 . The ECB cluster of genes defines a crucial cell cycle window during which a cell may change its fate; yet how the regulators that appear to act at ECBs are linked to cell cycle position is unclear , and coregulators , which experience tells us must exist , were unknown . Here , we describe our discovery that Bck2 , a potent cell-cycle-regulator whose function has remained obscure , functions as a cofactor for Mcm1 , to induce ECB–dependent gene expression . We also show that Bck2 has a role in promoting expression of late G1 and M/G2 genes . Our genetic and biochemical experiments reveal a new pathway for regulating gene expression associated with early cell cycle commitment , a process that is highly conserved .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"cell",
"division",
"gene",
"regulation",
"genetics",
"gene",
"expression",
"molecular",
"genetics",
"biology",
"cyclins",
"molecular",
"cell",
"biology",
"chromatin",
"dna",
"transcription",
"gene",
"function"
] |
2013
|
Bck2 Acts through the MADS Box Protein Mcm1 to Activate Cell-Cycle-Regulated Genes in Budding Yeast
|
Pathway analyses of genome-wide association studies aggregate information over sets of related genes , such as genes in common pathways , to identify gene sets that are enriched for variants associated with disease . We develop a model-based approach to pathway analysis , and apply this approach to data from the Wellcome Trust Case Control Consortium ( WTCCC ) studies . Our method offers several benefits over existing approaches . First , our method not only interrogates pathways for enrichment of disease associations , but also estimates the level of enrichment , which yields a coherent way to promote variants in enriched pathways , enhancing discovery of genes underlying disease . Second , our approach allows for multiple enriched pathways , a feature that leads to novel findings in two diseases where the major histocompatibility complex ( MHC ) is a major determinant of disease susceptibility . Third , by modeling disease as the combined effect of multiple markers , our method automatically accounts for linkage disequilibrium among variants . Interrogation of pathways from eight pathway databases yields strong support for enriched pathways , indicating links between Crohn's disease ( CD ) and cytokine-driven networks that modulate immune responses; between rheumatoid arthritis ( RA ) and “Measles” pathway genes involved in immune responses triggered by measles infection; and between type 1 diabetes ( T1D ) and IL2-mediated signaling genes . Prioritizing variants in these enriched pathways yields many additional putative disease associations compared to analyses without enrichment . For CD and RA , 7 of 8 additional non-MHC associations are corroborated by other studies , providing validation for our approach . For T1D , prioritization of IL-2 signaling genes yields strong evidence for 7 additional non-MHC candidate disease loci , as well as suggestive evidence for several more . Of the 7 strongest associations , 4 are validated by other studies , and 3 ( near IL-2 signaling genes RAF1 , MAPK14 , and FYN ) constitute novel putative T1D loci for further study .
Our approach builds on previous work that casts simultaneous analysis of genetic variants as a variable selection problem—the problem of deciding which variables ( the genetic variants ) to include in a multivariate regression of the phenotype . We begin with a method that assumes each variant is equally likely to be associated with the phenotype [75] , [76] , then we modify this assumption to allow for enrichment of associated variants in a pathway . The data from the GWAS are the genotypes and phenotypes from n study participants . We assume the genetic markers are single nucleotide polymorphisms ( SNPs ) , and the phenotype is disease status: patients with the disease ( “cases” ) are labeled , and disease-free individuals ( “controls” ) are labeled . Entries of the matrix are observed minor allele counts , or expectations of these counts estimated using genotype imputation [77] , [78] , for each of the n samples and p SNPs . We assume an additive model of disease risk , in which the log-odds for disease is a linear combination of the minor allele counts: ( 1 ) Under this additive model , is the odds ratio , the multiplicative increase in odds of disease for each copy of the minor allele at locus j . We do not consider dominant or recessive effects on disease risk , but it would be straightforward to include them; see [79] . This method is also easily adapted to quantitative traits by replacing ( 1 ) with a linear regression for y . Although the log-odds for disease is expressed in ( 1 ) as a linear combination of all SNPs , our framework is guided by the assumption that most SNPs have no effect on disease risk ( ) . While there is some debate over this assumption [80] , an advantage of this choice is that a SNP “included” in the multi-marker disease model—that is , a SNP j that has a non-zero coefficient , —indicates that the SNP is relevant to disease , or that it is in linkage disequilibrium with other , possibly untyped , variants that contribute to disease risk . Therefore , the main goal of the analysis is to identify regions of the genome that contain SNPs included in the disease model with high posterior probability , or identify SNPs within these regions that have a high “posterior inclusion probability , ” . A high PIP is the analogue of a small p-value in a conventional single-marker analysis . To obtain these posterior probabilities , we must first specify a prior for the coefficients . A standard assumption , and the assumption made in previous approaches [75] , [76] , is that all SNPs are equally likely to be associated with the phenotype a priori; that is , is the same for all SNPs . To model enrichment of associations within a pathway , we modify this prior . Precisely , the prior inclusion probability for SNP j depends on whether or not it is assigned to the enriched pathway: ( 2 ) where the pathway indicators record which SNPs are assigned to the enriched pathway; when SNP j is assigned to the enriched pathway , otherwise . ( In brief , a SNP is assigned to a pathway if it is near a gene in the pathway; see Methods . ) We refer to as the genome-wide log-odds , since it reflects the background proportion of SNPs that are included in the multi-marker disease model . ( More precisely , it is the proportion corresponding to SNPs not assigned to the pathway , which is usually most SNPs . ) We refer to as the log-fold enrichment because it corresponds to the increase in probability , on the log-odds scale , that a SNP assigned to the pathway is included in the model . For example , and indicates that 1 out of every 10 , 000 SNPs outside the pathway is included in the multi-marker model , but for SNPs assigned to the pathway , 1 out of every 100 is included . If , this reduces to the standard prior assumption made by previous methods . We expect to be 0 , or close to 0 , for most pathways . We assess enrichment by framing each hypothesis test for enrichment as a model comparison problem . To weigh the evidence for the hypothesis that candidate pathway with indicators is enriched for disease associations , we evaluate a Bayes factor [81] , [82]: ( 3 ) This Bayes factor ( BF ) is the ratio of likelihoods under two models , the model in which the candidate pathway is enriched for SNPs included in the multi-marker model ( ) , and the null model with no enrichment ( ) . A larger BF implies stronger evidence for enrichment . We compute each BF by averaging , or integrating , over the unknown parameters , and over multi-marker models with different combinations of SNPs , employing appropriate prior distributions for , , and the coefficients ( see Methods ) . Note that the Bayes factor ( 3 ) does not allow for a negative —that is , pathways that are underrepresented for associations with the phenotype . While it could be useful to investigate negative log-enrichments in other settings , in most GWAS of complex disease where there are generally few significant associations to begin with , reduced rates of disease associations in pathways would be difficult to find , and would be unlikely to have a useful interpretation . We use the same approach to test for joint enrichment of multiple candidate pathways . We compute as before ( eq . 3 ) , except that we set to 1 whenever SNP j is assigned to at least one of the enriched pathways . In this case , represents the increased level of associations ( on the log-odds scale ) among SNPs assigned to one or more of the pathways . This is equivalent to assuming that all enriched pathways have the same level of enrichment , which greatly simplifies the analysis . We allow for different enrichment levels only when accounting for enrichment of the MHC in RA and T1D . In that case , we have good reason to treat the MHC differently , given the predominant contribution of MHC alleles to RA and T1D risk [83] , [84] . To assess evidence for association of individual variants with the phenotype , we compute for each variant j . These posterior probabilities depend on which pathways are enriched , and on the log-fold enrichment , because these factors affect the prior probabilities , which in turn affect the posterior probabilities , following Bayes' rule . ( In practice , we account for uncertainty in and when calculating the posterior probabilities by averaging over and ; see Methods . ) Since enrichment leads to higher prior inclusion probabilities for SNPs in the enriched pathway , an association that is not identified by a conventional genome-wide analysis may become a strong candidate in light of its presence in an enriched pathway . Because we estimate from the data , the extent to which we prioritize variants is determined by the data . In this way , our framework integrates the problem of identifying enriched pathways with the problem of prioritizing variants near genes in enriched pathways .
We assemble a comprehensive list of candidate pathways to test for enrichment , drawing from a variety of publicly accessible collections ( see Methods ) . We do not filter pathway candidates based on their potential relevance to disease . In total , we interrogate 3158 candidate pathways for each disease , plus the MHC and ×MHC gene sets described below . Most candidate pathways were curated by domain experts , and others are based on experimental evidence in non-human organisms and inferred via gene homology . For full details of pathway databases used , and steps taken to compile gene sets from pathway data , refer to Methods , Supplementary Materials , and links to source code implementing our analyses . In two of the seven diseases , RA and T1D , multiple disease associations map to the major histocompatibility complex ( MHC ) region on chromosome 6 . Consequently , pathway analyses for RA and T1D tend to highlight pathways that involve MHC genes . When we apply our method to these diseases , the top pathways for T1D and RA are “Allograft rejection” and “Asthma , ” respectively . Both gene sets include multiple MHC genes , and exhibit strong evidence for enrichment ( ) . Other pathways with the strongest enrichment signals also contain MHC genes . Most of the support for enrichment of these pathways is likely driven by disease associations that map to the MHC . To check this , we create a “pathway” containing all genes within the MHC [90] , and test this gene set for enrichment . The MHC gene set shows more support for enrichment than any other pathway by several orders of magnitude ( in T1D and RA , respectively ) , and it is accompanied by a high enrichment estimate ( and 3 . 7 ) . Performing a similar enrichment analysis for all genes within the “extended” MHC ( xMHC ) [91] yields smaller BFs ( Figure S1 ) , suggesting that the genetic contribution to RA and T1D risk lies mostly within the class I , II and III subregions of the MHC . Our finding that the MHC is enriched for associations with T1D and RA is unsurprising considering the MHC is estimated to account for over half the genetic contribution to T1D risk , and at least a third for RA [83] , [84] , [92] . ( By contrast , the genetic contribution of the MHC is estimated to be ∼10% for CD [83] , and the BFs for the MHC and ×MHC in CD are 8 and 4 , respectively . ) In light of these findings , a reasonable question to ask is whether pathways show enrichment for disease associations beyond enrichment of the MHC . A strength of our model-based approach is that it can address this question by computing a BF for enrichment of each candidate pathway , conditioned on the estimated enrichment of the MHC . Thus , in our subsequent analysis of RA and T1D , we account for enrichment of disease associations within the MHC in this way . As far as we are aware , no other pathway-based analyses of these data incorporate enrichment of the MHC , which may explain why previous studies have highlighted mostly MHC-related pathways and gene categories . To give an initial impression of our enrichment results , we show the pathway with the largest BF for each disease in Figure 1 . The seven diseases exhibit a wide range of support for the strongest enrichment signal . For example , the top pathway for T1D , IL-2 signaling , has a BF of , whereas the largest BF for HT is only 5 . To address whether these top pathways constitute “significant” evidence for enrichment , the BF for enrichment must be weighed against the prior probability of the pathway being enriched to obtain a posterior probability of enrichment ( see “Interpretation of Bayes factors” in Methods ) . While specification of a prior probability of enrichment is subjective , this subjectivity is unavoidable; similar issues arise when specifying significance thresholds for p-values , though these issues are usually hidden ( 0 . 05 is a common threshold , but it is subjective and not universally appropriate [93] ) . If we apply a “conservative” value of 1/3158 to the prior probability for all candidate pathways , so that one pathway is expected to be enriched among the 3158 candidates , then CD and T1D show compelling evidence for enrichment ( posterior probability>0 . 99 ) , and RA shows suggestive evidence ( posterior probability = 0 . 45 ) . Considering the plausible connection between Measles pathway genes and RA ( discussed below ) , we view this as a compelling enrichment as well . The top pathway for T2D , Incretin regulation , shows only modest evidence for enrichment if we apply the conservative prior , but it might be considered “significant” if we adopt a less conservative prior to account for the known connection of this pathway to insulin resistance and diabetes . Based on these results , we do not investigate BD , CAD and HT further , and focus on the four diseases showing strongest evidence for enrichment , CD , RA , T1D and T2D . Figure 2 shows an expanded list of pathways with the strongest support for enrichment in CD , RA , T1D and T2D , together with estimated enrichment levels ( see Figure S1 for a longer list ) . Beyond these top results , the vast majority of candidate pathways show little or no evidence for enrichment ( Figure S2 ) , demonstrating that the method is robust to inclusion of many pathways that are most likely irrelevant to the disease . Before discussing the biological relevance of these pathways , we point out three general features of Figure 2 . First , some of the estimated enrichments are extremely large; for example , IL-2 signaling genes show more than a 10 , 000-fold enrichment of T1D risk factors . In contrast , the top pathway for CD , “Cytokine signaling in immune system , ” has roughly a 100-fold enrichment . ( Enrichment of this pathway nonetheless yields a large BF , partly because it implicates over 6700 SNPs; the BFs depend not only on the level of enrichment , but also on the number of SNPs assigned to the pathway . ) Second , some of the top pathways overlap or are subsets of one another . For example , “Cytokine signaling in immune system” is a subset of “Immune system . ” Also , the Immune system pathway from NCBI BioSystems ( BS ) overlaps with the Pathway Commons ( PC ) version of the same pathway ( 510 genes are common to both gene sets ) . This raises the question whether enrichment of just one pathway would suffice to explain the genome-wide association signal; we use our methods to investigate this question below . Third , 5 different pathway databases are represented in Figure 1 , and all 8 pathway databases included in our analysis appear among the top pathways ( Figure 2 ) , illustrating the benefits of interrogating pathways from multiple sources . The top-ranked pathway for CD ( “Cytokine signaling” ) is a collection of cytokine-driven networks that exhibit a complex relationship to autoimmunity—they promote inflammatory and immune responses , while also playing an important role in suppressing immunity [94] . Cytokine signaling implicates a broad class of 225 genes , suggesting that a collection of related gene networks explains the pattern of genetic associations better than any one signal transduction pathway . Enrichment of cytokine signaling is consistent with the accumulating evidence that points to cytokines , and the signaling cascades initiated by these cytokines , in a range of autoimmune disorders , including inflammatory bowel disease [9] , [95] , [96] . Previous findings from GWAS have linked autophagy genes ATG16L1 and IRGM to CD [65] , [97] , [98] . Our pathway analysis does not provide additional support for autophagy in CD because pathways reflecting current models of autophagy [9] , [99] have not yet been incorporated , as far as we are aware , into any of the publicly available pathway databases . Once we account for enrichment of the MHC , the top pathway for T1D is IL-2 signaling . Cytokine IL-2 and its interacting partners are indispensable to activation , development and maintenance of T regulatory cells , and disruption of IL2-mediated pathways promotes progression of autoimmune disorders [100]–[102] . T1D treatments targeting the IL-2 signaling pathway are currently undergoing clinical trials [103] . Additionally , studies in non-obese diabetic ( NOD ) mice suggest that defects in IL-2 signaling induce susceptibility to T1D [102] , [104] , [105] . Our findings support this hypothesis . The top pathway for RA , the KEGG “Measles” pathway , contains genes involved in immune response cascades triggered by infection of measles virus , including the cellular receptors expressed for measles virus such as SLAM and CD46 [106]–[108] . ( This result is again conditioned on enrichment of the MHC . ) While studies have associated the measles virus with RA [109] , other viral and bacterial infections have also been linked to incidence of RA [110] , [111] , and enrichment of the Measles pathway could reflect a larger class of genes involved in regulation of immune function during infection , rather than the measles virus specifically . The large BF for this pathway in both RA and T1D supports previous indications of a shared genetic basis [96] , [112] , and is consistent with observations that RA and T1D , along with other autoimmune diseases , recur in the same families [113] . All CD , RA and T1D pathways in Figure 2 implicate key actors in responses to pro-inflammatory stimuli and in regulation of innate and adaptive immunity . These include members of the NF-kB/Rel family , T-cell receptors ( TCRs ) , members of the protein tyrosine phosphatase family ( PTPs ) , mitogen-activated protein ( MAP ) kinases such as c-Jun NH2-terminal kinases ( JNKs ) , and chemokine receptors ( CXCRs ) [114]–[117] . None of the top-ranked pathways for CD , RA and T1D in our analysis have been identified in previous pathway-based analyses of these diseases [11] , [13] , . An important difference between our methods and previous pathway analyses of RA and T1D is that we incorporate enrichment of the MHC into models of enrichment . A previous analysis of RA [67] highlighted pathways “Bystander B cell activation” ( BioCarta , their p-value = ) and “Type 1 diabetes mellitus” ( KEGG , p-value = ) . However , both these pathways contain MHC genes , and our results suggest that enrichment of the MHC offers a better explanation of the association signal; in our analysis , support for enrichment of these pathways is several orders of magnitude less than support for enrichment of the MHC ( versus ) , and the support vanishes once we account for enrichment of the MHC ( BF = 0 . 69 , 0 . 57 ) . Similarly , previous analyses of T1D [16] , [121] have highlighted the same “Type 1 diabetes mellitus” pathway , but again support for enrichment is driven mostly by the association signal in the MHC , as our methods yield only modest support for this pathway after accounting for MHC enrichment ( BF = 43 ) . It is also notable that the top-ranked pathways for RA and T1D , Measles and IL-2 signaling , show strong support in our analysis only after accounting for enrichment of disease associations within the MHC; the BFs without MHC enrichment are 104 and 11 , whereas the BFs are and after conditioning on enrichment of the MHC . This may explain why these pathways have not been identified in previous pathway analyses of these diseases . These results illustrate the benefits of estimating enrichment conditioned on the MHC and , more generally , quantifying support for models with multiple enriched pathways . Another aspect that differs between our results and previous studies is that we interrogate a more comprehensive set of pathway databases . This may explain in part why the BF for the top pathway in CD , “Cytokine signaling in immune system” from Reactome , eclipses the BFs corresponding to previously reported pathways . For example , Wang et al [73] interrogated BioCarta , KEGG and Gene Ontology [46] ( and not Reactome ) gene sets for enrichment of CD associations , and reported the smallest p-value for BioCarta pathway “IL12 and Stat4 dependent signaling in Th1 development” ( p-value = , FDR = 0 . 045 ) . This pathway showed little evidence for enrichment in our analysis ( BF = 20 ) compared to cytokine signaling ( ) . ( Below , when we combine this pathway with cytokine signaling genes , we obtain stronger evidence for enrichment in CD; the BF is 81% the size of the largest BF for 2 enriched pathways . ) An important feature of our model-based approach is that pathway enrichments can help to map additional disease associations by prioritizing variants within enriched pathways . This is particularly useful for broad groups of enriched genes such as “Cytokine signaling in immune system , ” which contains 225 genes , as only a small portion of these genes may actually harbour genetic variants that affect CD risk . Prioritization occurs automatically within our statistical framework; the enrichment parameter affects the prior probability of association for SNPs in the pathway , which in turn increases the posterior probability of association for these SNPs . We therefore examine how re-interrogation of SNPs for association in light of inferred enrichments in CD , RA , T1D and T2D can reveal additional associations across the genome . We assess evidence for associations across genomic regions , rather than individual SNPs . The rationale is that genome-wide mapping using a multi-marker disease model sometimes spreads the association signal across nearby SNPs when they are correlated with one another , thereby diluting the signal at any given SNP [76] . We divide the genome into overlapping segments of 50 SNPs , with an overlap of 25 SNPs between neighbouring segments . For each segment , we compute , the posterior probability that at least one SNP in the segment is included in the multi-marker disease model . ( denotes the posterior probability that at least n SNPs are included . ) We use segments with an equal number of SNPs so that , under the null hypothesis of no enrichment , the prior probability that at least one SNP is included is the same for each segment . A segment spans , on average , 307 kb ( 98% of the segments are between 100 kb and 1 Mb long ) , so calculating for these segments provides only a low-resolution map of genetic risk factors for disease . Still , this resolution suffices for our study . Table 1 summarizes the regions of the genome showing strongest evidence for association after pathway prioritization , and Figure 3 compares support for disease associations under the null hypothesis of no enrichment with support under the model in which the pathway with the largest BF is enriched . Overall , prioritization of SNPs in enriched pathways increases support for disease risk factors in many regions , often substantially; these regions correspond to points above the diagonal in the scatterplots ( Figure 3 ) . In CD and RA , 8 disease susceptibility loci with , not including segments overlapping the MHC , are revealed only after prioritizing SNPs in enriched pathways . In T1D , prioritization of SNPs in the IL-2 pathway yields a total of 37 associated regions outside the MHC with . This dramatic result reflects the high estimated enrichment for IL-2 signaling genes . The majority of the additional disease regions with the strongest support , including many of the loci with weaker association signals , are validated by other studies; in CD and RA , 7 of the 8 additional disease susceptibility loci with are corroborated by other GWAS and large-scale meta-analyses , and in T1D , 4 of the 7 additional disease regions with are similarly corroborated ( see Table 1 for references ) . Prioritization yields many new candidate disease susceptibility loci not previously implicated by GWAS . These loci will require followup studies to be validated . Three unconfirmed T1D susceptibility loci with strong support ( ) are regions containing IL-2 signaling genes RAF1 , MAPK14 and FYN: gene RAF1 is a critical target of insulin in primary β-cells , and variants of this gene may modulate loss of β-cell mass in forms of diabetes [122]; MAPK14 ( p38 ) encodes at least 4 distinct isoforms , and deficiencies in one isoform have been shown in mice knockout studies to improve glucose tolerance and protect against insulin resistance , pointing to a role in development of T1D [123]; FYN interacts with PTPN22 to regulate T cell receptor signaling , and PTPN22 alleles are strongly associated with predisposition to T1D [124] . The one novel candidate region for RA contains Measles pathway gene TP73 , whose homolog , TP53 , is suspected to impair regulation of inflammation in RA patients [125] , [126] . Finally , conditioning on enrichment of top pathways in T2D yields a single novel candidate region at 7q32 with moderate probability of containing a disease association ( ) . This region contains GPR120 ( also OSFAR1 ) , a gene assigned to both the Incretin regulation and GLP-1 pathways . It was recently shown that GPR120-deficient mice develop obesity and reduced insulin signaling , and GPR120 expression is significantly higher in obese humans [127] , so the effect of this gene may be similar to the reported T2D association with FTO , in which variants near FTO increase T2D risk through an effect on body weight [128] , [129] . In addition to these associated regions , several promoted regions also lie within the MHC . ( In the scatterplots for RA and T1D , these regions correspond to open circles above the diagonal . ) Given the complexity of this region , which contains a high density of genes , and long-range correlations between SNPs , disentangling the association signal in the MHC will likely require higher density SNP data , and lies beyond the capacities of our current implementation ( see Discussion ) . Figure 4 compares the effect sizes of variants in regions selected only after accounting for pathway enrichment to the effect sizes from regions identified without the benefit of feedback from pathway enrichment . As expected , pathway prioritization uncovers many disease-associated variants with smaller effects than we would otherwise be able to map reliably . This could explain , at least in part , why many of the putative T1D associations uncovered in our analysis are not yet confirmed; the largest meta-analysis of T1D to date , with a combined sample of size ∼16 , 000 [130] , still has limited power to detect associations within this range of effect sizes and minor allele frequencies . ( In contrast , much larger meta-analyses exist for CD and RA , with 30 , 000 and 47 , 000 samples , respectively [131] , [132] . ) Finally , Figure 3 offers the opportunity to remark on four other features of our results . First , the strongest association signals without feedback from pathways stay strong whether or not they are related to the enriched pathway—these associations correspond to the points in the top-right corner of each scatterplot . ( Note that the segments in the top-right corner recapitulate the strongest associations reported for CD , RA , T1D and T2D in the original study [65] . See Supplementary Materials for a detailed comparison to single-marker p-values in all 7 diseases . ) Second , many segments show slightly decreased support for association under the enrichment hypothesis ( points below the diagonal in the scatterplots ) . This occurs because the estimated prior inclusion probability for SNPs outside the pathway is reduced to reflect the fact that pathway enrichment helps to explain an appreciable portion of the genome-wide association signal . Third , although not evident from the figure due to over-plotting , most segments show little or no evidence for associations under either hypothesis; in each scatterplot , 98–99 . 7% of segments lie near the bottom-left corner . Fourth , associations with strong support under the null are not necessary for establishing evidence for enriched pathways; none of the RA associations in the top-right corner of the scatterplot contribute to evidence for enrichment of the Measles pathway . Above , we obtained evidence for enriched pathways in CD , RA and T1D . The question remains whether a combination of several enriched pathways offers a better fit to the data . A benefit of our approach is that we can compare support for enrichment of different combinations of pathways by comparing their BFs ( assuming the same prior for these enrichment hypotheses ) . We assess support for combinations of pathways in CD , RA and T1D by computing BFs for models in which 2 and 3 pathways are enriched . Since it is impractical to consider all combinations of 2 and 3 pathways , we tackle this in a “greedy” fashion by selecting combinations of pathways based on the initial ranking ( see Methods ) . Figure 5 gives the combinations of 2 and 3 pathways that yield the largest BFs for these diseases . Again , to properly interpret these results we must weigh these BF gains against the relative prior plausibility of the models . Using a “conservative” prior for any pair of pathways being enriched ( see “Interpretation of Bayes factors” in Methods ) , we interpret Figure 5 as providing considerable , if short of compelling , support for the hypothesis that 2 pathways are enriched for disease associations in CD , RA and T1D . For example , in CD the BF for enrichment of both cytokine signaling and IL-23 signaling genes is 377 times greater than the BF for enrichment of cytokine signaling genes alone . The BFs for models in which 3 pathways show further increases , but not enough to constitute strong evidence for enrichment of 3 pathways . We also examine whether enrichment of multiple pathways can lead to identification of additional loci affecting susceptibility to disease . Figure S3 shows that allowing for 2 enriched pathways in CD , RA and T1D does not yield strong support for associations beyond what is already revealed by enrichment of the single top pathway . We do , however , find that a segment near IL12B shows a substantial gain in support for association with CD ( increases from 0 . 03 to 0 . 44 ) , and this association is confirmed by other GWAS [6] , [7] , [132] , [133] .
Motivated by the observation that it is easier , at least in principle , to identify associations within an enriched pathway , we developed a data-driven approach to simultaneously assess support for enrichment of disease associations in pathways , and prioritize variants in enriched pathways . We investigated the merits and limitations of this approach in a detailed analysis of data sets for seven complex diseases . We interrogated thousands of candidate pathways from multiple pathway databases , finding strong evidence linking pathways to pathogenesis of several diseases . By promoting variants within the enriched pathways identified in our analysis , we mapped disease susceptibility loci beyond those identified by a conventional analysis . The CD and RA results provided some validation for our methods , as all but one of the additional disease associations identified by pathway prioritization are corroborated by other studies . The T1D results also provided some validation for our methods , as several of the strongest associations informed by enrichment of the IL-2 signaling pathway are confirmed in other GWAS for T1D . Prioritizing IL-2 signaling genes revealed other regions relevant to T1D that could not be corroborated by other GWAS , and this may be because the largest GWAS for T1D to date does not match the scale of the largest studies for CD and RA . All the disease associations informed by enriched pathways had smaller effects on disease susceptibility , illustrating how pathway prioritization can help overcome some of the constraints on our ability to reliably detect disease-conferring variants with small effects in GWAS . Our approach builds on methods that use multi-marker models to simultaneously map associated variants in GWAS [61] , [75] , [76] , [134]–[142] . In contrast to single-marker regression approaches , these methods model susceptibility to disease by the combined effect of multiple variants , and use sparse multivariate regression techniques to fit multi-marker ( i . e . polygenic ) models to the data . Within a multi-marker model of disease , estimating enrichment of a candidate pathway effectively reduces to counting , inside and outside the pathway , variants associated with disease ( more precisely , variants included in the multi-marker model ) . Our approach to combining multi-marker modeling with pathway analysis offers several benefits . First , unlike many pathway analysis methods that test for enrichment of significant SNPs or genes within a pathway [24] , [25] , we have no need to select a threshold to determine which p-values are significant; instead , we use the association signal from all variants to assess enrichment . Second , by analyzing variants simultaneously , we avoid exaggerating evidence for enrichment from disease-associated variants that are correlated with each other ( i . e . in linkage disequilibrium ) , while still allowing multiple independent association signals near a gene to contribute evidence for enrichment . Third , and most importantly , quantifying enrichment within this framework gives us feedback about associations within enriched pathways , potentially leading to discovery of novel genetic loci underlying disease . In contrast to many pathway analysis methods , we modeled enrichment of disease associations at the level of variants , rather than genes . While there are arguments for both approaches , a feature of the variant-based approach is that , when there are multiple variants near a gene that affect disease susceptibility , all these signals contribute to the evidence for enrichment of pathways containing this gene . Another important feature of our approach is that it can be used to assess models in which multiple pathways are enriched . Examining combinations of pathways for enrichment may highlight pathways that would otherwise not be highly ranked . The results on RA and T1D provided vivid examples of this; evidence for enrichment of the Measles and IL-2 pathways only became compelling once we assessed support for enrichment of these pathways together with enrichment of the MHC . Our results focused on the regions showing the strongest evidence for association with disease . However , the large number of points approaching the middle of the vertical axes in Figure 3 suggests that many other gene variants in the enriched pathways may contribute to risk of CD , RA and T1D; from our estimates of and ( Figure 2 ) , approximately 38 , 45 and 59 independent risk variants are , in expectation , hidden among Measles , cytokine signaling and IL-2 signaling genes , respectively . This suggests that more disease associations in these pathways remain to be discovered . Several selected disease susceptibility regions ( Table 1 ) contain multiple candidate genes , including cases in which the gene in the enriched pathway is not the same as the most credible gene suggested in prior studies . It is possible that pathway annotations would be useful to help pinpoint , or fine-map , the genes or variants relevant to disease within these regions . However , investigating this would require advances to our current methodology , as the approximations we made to improve the efficiency of our approach , building on earlier work [75] , are less appropriate for refining the location of association signals , and these approximations will need to be modified to accommodate this aim . Nonetheless , we note that some of these regions may contain multiple variants that disrupt or regulate genes relevant to disease , and our methods can help assess this possibility . For example , we calculate that multiple independent risk variants reside at the 16p13 locus with probability ( Table 1 ) . So it is possible that both C1QTNF6 and IL2RB at this locus are associated with T1D risk variants . A limitation of our current approach is that the prior variance of additive effects on disease risk must be chosen beforehand . We based our choice on the distribution of odds ratios reported in published genome-wide association studies , and checked that the ranking of enriched pathways was robust to different prior choices ( see Methods ) . One problem with this prior is that published associations typically have the largest effects on disease risk , as these are the associations we usually have adequate power to identify . This results in a prior that places too much weight on larger additive effects . It would be preferable to estimate this prior from the data instead , but we found that this worked poorly in practice . The likely cause of this problem is that the non-zero effects on disease are not normally distributed , contrary to our assumptions . One possible solution would be to use a more flexible prior that is better able to capture the distribution of additive effects , such as a mixture of two or more normals [80] . In summary , our results on a range of complex diseases illustrate how an integrated approach to identification of enriched pathways , and prioritization of variants within enriched pathways , can identify additional disease associations beyond standard statistical procedures based on single-marker regression . Our results point to the potential for applying our methods to other common diseases , and larger studies , to uncover genetic loci that have not yet been identified as risk factors for disease .
Results on all seven diseases are based on genome-wide marker data from the case-control studies described in the original WTCCC study [65] . For all diseases , the control samples come from two groups: 1480 individuals from the 1958 Birth Cohort ( 58BC ) , and 1458 individuals from the UK Blood Services ( UKBS ) cohort . All subjects are from Great Britain , and are of self-described European descent . Genetic associations from these studies were first reported in [65] . All study subjects were genotyped for roughly 500 , 000 SNPs on autosomal chromosomes using a commercial version of the Affymetrix GeneChip 500K platform . We estimate missing genotypes at the SNPs using the mean posterior minor allele count from BIMBAM [79] , [143] , with SNP data from Phase II of the International HapMap Consortium project [144] . To be consistent with the original analysis , refSNP identifiers and locations of SNPs are based on Human Genome Reference Assembly 17 ( NCBI build 35 ) . We apply quality control filters as described in [65] , and remove SNPs that exhibit no variation in the sample . For all diseases , we include an additional quality control measure to filter out potentially problematic SNPs . Some SNPs with high minor allele frequencies ( MAFs ) show moderate evidence for association based on our calculations—“single-SNP” BFs [79] in which the prior standard deviation of the log-odds ratios is set to 0 . 1—but because they do not appear to be supported by nearby SNPs upon inspecting their single-SNP BFs , we cannot rule out the possibility of genotyping errors . Based on this criterion , we discard 2 additional SNPs in CD , rs1914328 on chromosome 8 at 69 . 45 Mb ( , MAF = 0 . 43 ) , and rs6601764 on chromosome 10 at 3 . 85 Mb ( , MAF = 0 . 43 ) . In each case , no nearby SNPs have single-SNP BFs greater than 46 . For CAD , we discard SNP rs6553488 on chromosome 4 at 171 . 4 Mb ( , MAF = 0 . 46 ) . No nearby SNPs have a single-SNP BF greater than 11 . Following the same quality control criterion , we do not filter out SNPs in the other data sets . Table S1 summarizes the data used in our analysis after following these quality control steps . We aim for a comprehensive evaluation of pathways accessible on the Web in standard , computer-readable formats [63] , [64] , [145] . Since the results hinge on the quality of the pathways used in the analysis , we restrict the analysis to curated , peer-reviewed pathways based on experimental evidence , and pathways inferred via gene homology . We draw candidate pathways from the collections listed in Figure 6 ( see also Supplementary Materials ) . KEGG [146] and HumanCyc [147] are primarily databases of metabolic pathways , and are unlikely to be relevant to some autoimmune diseases , but for completeness we include them in the analysis of all diseases . We create 2 additional gene sets to assess support for enrichment of disease associations within the MHC and “extended” MHC ( xMHC ) [90] , [91] . We treat each candidate pathway as a set of genes , ignoring details such as molecules involved in biochemical reactions , and cellular locations of these reactions . Many pathways are arranged hierarchically in the databases; we include all elements of the hierarchy in our analysis . Elements in upper levels of the hierarchy refer to groups of pathways with shared attributes , or a common function . Some gene groups have a broad definition , such as “Immune system” in Reactome ( ID 6900 ) , which includes pathways involved in adaptive and innate immune response . Enrichment of a large gene set is unlikely to provide much insight into disease pathogenesis . However , a key step in our analysis is to re-interrogate SNPs for association in light of inferred enrichments . Thus , enrichment of a broad physiological target such as “Immune system” can be useful if subsequent re-interrogation reveals associations that were not significant in a conventional analysis . Since we combine pathways from different sources , we encounter pathways with inconsistent definitions [148] , [149]; see Supplementary Materials . There is no single explanation for this lack of consensus , and we have no reason to prefer one definition over another , so we include all versions of pathways in our analysis . Based on findings that the majority of variants modulating gene expression lie within 100 kb of the gene's transcribed region [150]–[152] , we assign a SNP to a gene if it is within 100 kb of the transcribed region . Others have opted for a 20 kb window [69] , [73] based on findings that cis-acting expression QTLs are rarely more than 20 kb from the gene [62] . We choose a broader region since the benefit of including potentially relevant SNPs in a pathway when the association signal is sparse seems likely to outweigh the cost of including a larger number of irrelevant markers . In our case studies on CD , RA and T1D , we compute BFs to assess support for models in which 2 and 3 pathways are enriched for disease associations . Since it is impractical to consider all combinations of 2 and 3 pathways , we tackle this in a “greedy” fashion by selecting combinations of pathways based on the initial ranking . Our strategy is to select the pathway with the largest BF ( Figure 1 ) , and assess support for this pathway in combination with pathways from a larger set of candidates ( we take all pathways with BF>10 ) . This heuristic makes it feasible to evaluate many combinations of pathways that could plausibly be jointly enriched , though it does not consider all combinations , so we may miss a combination with stronger evidence for enrichment . In total , we compute BFs for 85 , 24 and 408 pairs of pathways in CD , RA and T1D , respectively . ( Note that models for RA and T1D also include enrichment of the MHC . ) For completeness , we extend the analysis to models with 3 enriched pathways . Following the same greedy strategy , we take the top pair of pathways ( Figure 5 ) and combine it with individual pathways with BF>10 . The Bayesian variable selection approach to simultaneous interrogation of SNPs involves fitting a multi-marker disease model to the data with different combinations of SNPs . By accounting for correlations between markers , fitting all markers simultaneously allows us to identify those that are independently associated—that is , markers that individually signal a variant contributing to disease risk independently of other risk-conferring variants . MATLAB implementations of the statistical methods described here , and the MATLAB scripts used to implement the steps in our analysis , are available for download at http://github . com/pcarbo/bmapathway .
|
Genome-wide association studies have helped locate gene variants that affect our susceptibility to diseases . The analysis of these studies is typically straightforward: test each genetic variant whether it is correlated with predisposition to disease . This approach often works well for identifying commonly occurring variants with moderate effects on disease risk . However , the effects of many variants are so small they fail to register statistically significant correlations . This is a concern because many diseases are modulated by many genetic factors with small effects on disease risk . An alternative is to examine groups of variants , such as variants sharing a common pathway , and assess whether these groups are “enriched” for correlations with disease . This can be a more effective approach to identifying genetic factors relevant to disease . However , it does not tell us which genes are associated with disease . To address this limitation , we describe an approach that integrates enrichment analysis with tests for disease-variant correlations within a single framework . We illustrate this approach in genome-wide studies of seven complex diseases . We show that our approach supports enriched pathways in several diseases , and uncovers disease-susceptibility genes in these pathways not identified in conventional analyses of the same data .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2013
|
Integrated Enrichment Analysis of Variants and Pathways in Genome-Wide Association Studies Indicates Central Role for IL-2 Signaling Genes in Type 1 Diabetes, and Cytokine Signaling Genes in Crohn's Disease
|
The phytohormone abscisic acid ( ABA ) is critical to plant development and stress responses . Abiotic stress triggers an ABA signal transduction cascade , which is comprised of the core components PYL/RCAR ABA receptors , PP2C-type protein phosphatases , and protein kinases . Small GTPases of the ROP/RAC family act as negative regulators of ABA signal transduction . However , the mechanisms by which ABA controls the behavior of ROP/RACs have remained unclear . Here , we show that an Arabidopsis guanine nucleotide exchange factor protein RopGEF1 is rapidly sequestered to intracellular particles in response to ABA . GFP-RopGEF1 is sequestered via the endosome-prevacuolar compartment pathway and is degraded . RopGEF1 directly interacts with several clade A PP2C protein phosphatases , including ABI1 . Interestingly , RopGEF1 undergoes constitutive degradation in pp2c quadruple abi1/abi2/hab1/pp2ca mutant plants , revealing that active PP2C protein phosphatases protect and stabilize RopGEF1 from ABA-mediated degradation . Interestingly , ABA-mediated degradation of RopGEF1 also plays an important role in ABA-mediated inhibition of lateral root growth . The presented findings point to a PP2C-RopGEF-ROP/RAC control loop model that is proposed to aid in shutting off ABA signal transduction , to counteract leaky ABA signal transduction caused by “monomeric” PYL/RCAR ABA receptors in the absence of stress , and facilitate signaling in response to ABA .
Abscisic acid ( ABA ) is a phytohormone that protects plants against abiotic stress and is involved in seedling development . In response to abiotic stress conditions , ABA concentrations rise in plant cells [1–3] . ABA can be perceived by a group of soluble “PYL/RCAR” ABA receptor proteins , which upon ABA binding lead to formation of ABA-PYL/RCAR-PP2C complexes that in turn inhibit PP2C protein phosphatase activity [4 , 5] . This releases PP2C-mediated inhibition of the downstream SnRK2 protein kinases [6 , 7] . Subsequently , SnRK2 protein kinases are activated and phosphorylate downstream transcription factors and ion channels to trigger ABA responses [1 , 3 , 8] . Recent structural and biochemical studies revealed that 14 Arabidopsis PYL/RCARs can be subdivided into two groups: those with a higher probability of forming PYL/RCAR dimers and those with a thermodynamically favored monomeric state . Monomeric PYL/RCARs bind to PP2Cs and may downregulate PP2C activity even in the absence of the ABA ligand [9 , 10] . Theoretically , this constitutive receptor activity would cause “leaky” ABA signal transduction [10] . Diverse receptor-mediated signal transduction mechanisms in eukaryotes include specific proteins that can shut off signaling in the absence of the stimulus , e . g . , [11–13] . However , molecular mechanisms that protect against leaky ABA signal transduction remain unknown . Reconstitution studies of ABA signal transduction have demonstrated the fundamental roles of a set of core ABA signaling components PYL/RCAR-PP2C-SnRK/CDPKs [14–17] . Studies suggest that additional factors function in ABA signal transduction [18–22] . The linkage of some factors with core ABA signal transduction components needs to be further addressed . RopGEFs are plant-specific “PRONE” ( plant-specific ROP nucleotide exchanger ) -domain-containing guanine nucleotide exchange factors [23 , 24] . The Arabidopsis genome encodes 14 RopGEFs with a high degree of sequence similarity , particularly for the residues that are involved in catalyzing GDP/GTP exchange [24] . Insights from the crystal structures of ROP-GDP-PRONE ternary intermediates and ROP-PRONE binary complexes revealed the molecular mechanism of activation of the small GTP-binding proteins in plants by RopGEFs through promoting exchange of GDP for GTP [25 , 26] . Besides activation of ROP/RACs , recent studies have suggested that RopGEFs also act as a bridge in the linkage of signal transduction from receptor-like kinases ( RLKs ) to ROP/RACs . The FERONIA RLK was shown to function as an upstream regulator of ROP/RAC signaling likely by interacting with RopGEFs to mediate auxin effects on root hair growth [27] . Another receptor-like kinase , AtPRK2 , physically binds to and phosphorylates RopGEF1 in the C terminal region , which in turn promotes ROP1 activation during pollen tube growth [28] . In previous studies , it was demonstrated that ROP/RAC small GTPases act as negative regulators of ABA signal transduction [29 , 30] , and ROP11 can directly bind to the type 2C protein phosphatase ABI1 and protect ABI1 protein phosphatase activity from inhibition by RCAR1/PYL9 [31] . Furthermore , ABA did not affect the mRNA or protein levels of ROP11 in plants [32] . However , the mechanisms by which ABA regulates activity of the ROP11 GTPase remain unknown . We investigated whether ABA may control the behavior of ROPs through affecting the ROP activators RopGEFs . Here , we show that ABA signal transduction directly mediates the removal of RopGEF1 via relocation to and degradation in vacuoles , thus enabling robust ABA signal transduction . Moreover , PP2C phosphatases directly interact with RopGEF1 and inhibit the degradation of RopGEF1 in the absence of ABA , thus providing a mechanism to ensure shutting off of ABA signal transduction . Based on our results , a GEF-ROP-PP2C control loop model is suggested that prohibits leaky ABA signal transduction and that functions in modulation of ABA signaling strength .
To study the effects of ABA signal transduction on RopGEF1 protein ( from here on “GEF1” ) , we first investigated the subcellular localization of GEF1 in the absence and presence of exogenous ABA . We expressed GFP-GEF1 , GEF1-GFP , and GEF1-mCherry fusion proteins in Nicotiana benthamiana epidermal cells . All three constructs showed similar fluorescence signal patterns ( S1A–S1C Fig ) . We then generated GFP-GEF1 overexpression lines in Arabidopsis wild-type plants . These transgenic lines exhibited swollen root hairs similar to those found upon overexpression of GEF1 only or overexpression of a constitutively active ROP11 ( S1F Fig ) [33] , a downstream substrate of GEF1 [34] . These observations indicated that the GFP-GEF1 construct is functional in Arabidopsis plants . Confocal microscopy revealed that in the absence of exogenous ABA , the GFP-GEF1 signal was largely located in the cytosol and cell periphery ( Fig 1A left and S1A–S1C Fig ) , where GEF1-mCherry fluorescence signal was visible following NaCl-induced plasmolysis ( S1C Fig ) and GFP-GEF1 fluorescence partially overlapped with the lipophilic dye FM4-64 signal , which stains the plasma membrane ( S1E Fig ) . Surprisingly , after application of ABA , numerous intracellular fluorescent particles appeared in Arabidopsis root epidermal cells ( Fig 1A right ) . Further analyses showed that formation of these particles was time- and ABA-dose-dependent ( Fig 1B and 1C ) . Experiments investigating different hormone stimuli including auxin , GA , ethylene , JA , and brassinosteroid demonstrated that this effect was specific to ABA relative to these hormones ( S2A Fig ) . To better visualize this ABA-mediated formation of intracellular particles , we examined the subcellular localization of GFP-GEF1 in N . benthamiana leaves treated with ABA . Time-course confocal microscopy revealed that within 5 min of ABA treatment , GFP-GEF1 moved from the cell periphery and gradually accumulated in numerous particles ( Fig 1D ) . Subsequently , the fluorescence intensity was substantially reduced over a time period of 30–180 min ( Fig 1B and 1D ) . In comparison , the subcellular localization and fluorescence signal intensity of GFP alone was not altered under the same ABA treatment ( Fig 1E ) . Furthermore , GFP-GEF1 driven by the GEF1 promoter also formed particles in response to ABA ( S2D Fig ) . In controls without ABA addition , 3 h of microscopy did not cause GFP-GEF1 to form particles ( S2E Fig ) . Experiments were carried out to determine the identity of the ABA-induced GFP-GEF1 particles . We investigated the co-localization of GFP-GEF1 in N . benthamiana leaves with mCherry labeled organelle markers [35] . No co-localization or sparse overlap was observed with mitochondrial , peroxisomal , ER , and cis-Golgi markers ( Figs 2A and S3A ) , indicating that GFP-GEF1 proteins were not sequestered into these endomembrane systems in response to ABA . In comparison , overlap of the GFP and mCherry signals was observed in co-localization analysis with the retromer marker SNX1 [36] and the prevacuolar compartment ( PVC ) markers BP-80 [37] and VPS45 ( Fig 2A ) [38] . Reminiscent of recent studies showing that the plasma membrane proteins , FLAGELLIN SENSITIVE2 ( FLS2 ) , the brassinosteroid receptor BRI1 , and the auxin efflux carrier PIN2 are internalized rapidly from the plasma membrane and sequentially delivered into the vacuole via the late endosome [39–41] , we speculated that GFP-GEF1 proteins were also subjected to this process in response to ABA . In line with this hypothesis , ABA-induced GFP-GEF1 particle formation was sensitive to Wortmannin ( Fig 2B ) , a phosphatidylphosphate-3-kinase ( PI3K ) inhibitor that interferes with vesicle trafficking from the plasma membrane to the prevacuolar compartment [42 , 43] . GFP-GEF1 particles accumulated into typical Wortmannin-induced intracellular ring-like structures ( Fig 2B ) [44 , 45] . In addition , we observed that ABA-induced GFP-GEF1 particles partially co-localized with an endosomal ( LE ) marker ARA7 ( S3A Fig ) [46] . In response to stimulation by a combination of ABA and Brefeldin A ( BFA ) , an inhibitor of endosomal transport [47] , GFP-GEF1 particles aggregated into large blocks ( S3C Fig ) , known as BFA bodies [48] . Furthermore , subcellular localization of GFP-GEF1 did not showed significant change in response to Wortmannin or BFA treatment , respectively ( S3C Fig ) . The above analysis suggested that GFP-GEF1 proteins are relocated to the prevacuolar compartment via an endosome–prevacuolar compartment pathway in response to ABA . Proteins in the prevacuolar compartment are further transported into vacuoles through membrane fusion and are subsequently subjected to either storage or degradation [49] . To trace the fate of GFP-GEF1 , we carried out western blot analyses to detect GFP-GEF1 fusion protein abundance in response to ABA in transgenic Arabidopsis plants . Immunoblot analyses revealed that GFP-GEF1 protein levels were rapidly and substantially reduced in response to ABA treatment ( Fig 3 ) , but only slightly altered in response to the control ethanol ( EtOH ) stimulus ( solvent used for ABA ) and combined ABA and Wortmannin ( Fig 3 ) . Moreover , ABA-mediated degradation of GFP-GEF1 was still evident in the presence of the proteasome inhibitor MG132 ( Fig 3 ) , suggesting no major role of the proteasome in this degradation response . Furthermore , confocal microscopic observations indicated that ABA-mediated GFP-GEF1 particles showed co-localization with the vacuolar marker γ-TIP ( S3B Fig ) [50 , 51] . Taken together , we concluded that GFP-GEF1 is sequestered to vacuoles and is degraded in response to ABA . We pursued yeast-two-hybrid ( Y2H ) experiments with GEF1 as bait in search of key ABA signal transduction factors that might participate in ABA-mediated degradation of GEF1 . The type 2C protein phosphatase ( PP2C ) ABI1 was identified as a candidate GEF1 interactor . Interaction between ABI1 and GEF1 was observed in Y2H assays ( Fig 4A ) . Known interactions between ABI1 and RCAR1/PYL9 [4] or GEF1 and ROP11 [34] served as positive controls ( Fig 4A ) . In comparison with ABI1 , other PP2Cs including ABI2 , HAB1 , and PP2CA exhibited weak interactions with GEF1 ( Fig 4A ) . However , no reproducible interactions were observed for GEF1 and OST1 , which is a direct substrate of ABI1 [6 , 52 , 53] . Lack of ABI1 and RCAR12/PYL1 interactions in the absence of ABA was used as a negative control ( Fig 4A ) . The putative interaction of ABI1 and GEF1 was further investigated in in vitro pull-down assays . Using GST-tagged GEF1 as bait , Strep-II tagged ABI1 and also ROP11 but not OST1 co-immunoprecipitated with GST-GEF1 ( Fig 4B ) . Next , we pursued bimolecular fluorescence complementation ( BiFC ) assays to investigate putative interactions between GEF1 and PP2Cs in plant cells . Surprisingly , numerous fluorescent particles were detected in the cytosol ( S4A Fig ) . Quantitative analyses of BiFC signals showed strong ABI1-GEF1 interactions and weak signals for GEF1 interaction with other clade A PP2Cs ( S4A and S4B Fig ) , which correlated with observations from Y2H experiments ( Fig 4A ) . In positive controls , strong BiFC signals for ROP11 and GEF1 interaction were observed , which appeared at the cell periphery ( S4 Fig ) . Co-localization analyses with GFP-GEF1 and mCherry-ABI1 indicated that GEF1 and ABI1 localizations overlapped at the cell periphery without ABA treatment ( S5A Fig ) . Furthermore , Y2H and Co-IP experiments showed that GEF1 also interacted with additional PP2Cs in the clade A PP2C phosphatase family , including HAI1 and AHG1 ( S5B and S5C Fig ) . We performed experiments to determine whether PP2Cs are involved in regulating the formation of intracellular GFP-GEF1 particles . We expressed the same pUBQ-GFP-GEF1 construct as used in wild-type Arabidopsis plants for subcellular localization analyses ( Fig 1A and 1B ) in the pp2c quadruple mutant background abi1/abi2/hab1/pp2ca and investigated the subcellular localization of GFP-GEF1 . Over 100 transgenic lines were checked , and all transgenic lines showed extremely weak GFP signals compared with those in the wild-type background ( Fig 5A ) . RT-PCR and quantitative RT-PCR analyses showed that this is not because of a reduced transcript level of the GEF1 mRNA ( Figs 5B and S8C ) . Interestingly , subcellular localization analyses of GFP-GEF1 showed fluorescence in intracellular particles of abi1/abi2/hab1/pp2ca root cells even without ABA treatment ( Fig 5A ) . To investigate the identity of these particles , transgenic GFP-GEF1/abi1abi2hab1pp2ca seedlings were subjected to Wortmannin treatment . Confocal microscopic observations indicated that these particles partially aggregated into ring-like structures 1 h after Wortmannin treatment ( S6A Fig ) . After prolonged Wortmannin exposure of GFP-GEF1/abi1abi2hab1pp2ca plants , GFP-GEF1 fluorescence signals appeared in the cell periphery and perinuclear region ( S6B Fig ) . This cellular localization pattern resembled that of subcellular localization GFP-GEF1 in the wild-type background in the absence of ABA ( S1A–S1C Fig ) . These confocal microscopic analyses strongly suggested that GFP-GEF1 protein in abi1/abi2/hab1/pp2ca plants underwent constitutive degradation in the absence of added ABA . These findings suggested that the wild-type PP2Cs prohibit spontaneous degradation of GFP-GEF1 . Furthermore , in newly emerging lateral roots , the intracellular particles were not degraded and instead aggregated into large intracellular fluorescent bodies ( S6C Fig ) . These observations are consistent with the lack of the expression of the vacuolar TIP markers in these developing new Arabidopsis root tip cells that have been proposed to lack lytic vacuoles at this developmental stage [54 , 55] . We next investigated GFP-GEF1 protein levels in GFP-GEF1/abi1abi2hab1pp2ca plants by western blotting . Although GEF1 mRNA levels were abundant in GFP-GEF1/abi1abi2hab1pp2ca plants ( Figs 5B and S8C ) , we could not detect a clear western blot signal , indicating the GFP-GEF1 protein levels are extremely low in the abi1/abi2/hab1/pp2ca quadruple mutant in contrast to wild-type plants ( Fig 5C ) . Additional western blot experiments showed that GFP-GEF1 protein levels were measurably enhanced after 3 h of Wortmannin treatment ( S6D and S6E Fig ) . Taken together , these data show that PP2C protein phosphatases interact with GEF1 ( Figs 4 , S4 and S5 ) , and in wild-type plants prevent GFP-GEF1 degradation in the absence of ABA ( Fig 5 ) . Previously , the small GTPase ROP11 in its active form was shown to directly bind to ABI1 and protect ABI1 protein phosphatase activity from inhibition by the ABA receptor RCAR1/PYL9 [31] . Given that GEFs activate ROPs through facilitating GDP/GTP exchange [25 , 26] , we speculated that GEF1 plays a positive role in PP2C function . To dissect the genetic relevance between GEF1 and the PP2C protein phosphatases , we analyzed ABA-mediated phenotypes in GFP-GEF1 /abi1abi2hab1pp2ca overexpression plants . We first examined ABA-mediated inhibition of seed germination and seedling establishment . After 3 d of stratification , abi1/abi2/hab1/pp2ca mutant seeds showed a late germination phenotype on 1/2 MS medium regardless of the presence of ABA ( Fig 6A and 6B ) . In comparison , three independent GFP-GEF1/abi1aib2hab1pp2ca transgenic lines showed partial but not full restoration of seed germination and seedling establishment in the pp2c quadruple mutant background ( Fig 6A–6F ) , which may be linked to the interaction of GEF1 with additional clade A PP2Cs ( S5B and S5C Fig ) . We then analyzed the responses of these genotypes in ABA-mediated inhibition of primary root elongation . To ensure simultaneous seed germination , seeds of all genotypes were stratified for 10 d . Seedlings with similar root lengths were transferred onto 1/2 MS plates supplemented with or without ABA . The results indicated that GFP-GEF1/abi1abi2hab1pp2ca seedlings were less sensitive to ABA-mediated inhibition of primary root growth than the pp2c quadruple mutant ( Fig 6E and 6F ) . These phenotypic assays provide initial evidence for a genetic interaction between GEF1 and PP2Cs in ABA responses . To pursue a more direct genetic investigation of GEF1 functions in ABA signaling , we examined ABA-related phenotypes in GEF1 single knockout mutant and GEF1 overexpression lines . However , we did not observe any differential responses to ABA- mediated inhibition of seed germination and primary root growth in these two genotypes compared with wild-type ( Fig 7A and 7B ) . We speculated that this result was attributable to possible overlapping gene functions , considering the high sequence homology among members of the large GEF family . To circumvent possible redundancy , we pursued a generation of higher-order gef mutants . Considering the rapid GEF1 degradation in response to ABA , we assumed that other GEFs with similar functions as GEF1 in ABA signal transduction should be removed in the same manner in response to ABA . Based on this hypothesis , we examined the subcellular localization of all 14 members of the GEF family in the absence and presence of ABA . The results showed that like GEF1 , GEF4 , 10 , 12 , and 14 formed particles in response to ABA ( Fig 7C ) . Other GEFs did not show ABA-induced intracellular particle formation ( e . g . , GEF7; Fig 7C ) . By analyzing ABA-mediated phenotypes in higher-order gef mutants , we found that a triple mutant gef1/4/10 showed a slightly enhanced sensitivity to ABA-mediated inhibition of primary root elongation and seed germination compared to wild-type ( Fig 7D and 7E ) , consistent with a previous report [56] . This slight hypersensitivity to ABA-mediated inhibition of seed germination was further enhanced in gef1/4/10/14 quadruple mutant seed ( Figs 7F and S11 ) . The above analyses provided evidence to support that GEFs act as negative regulators of ABA responses . During exploration of the biological significance of ABA-mediated degradation of the GEF1 protein , we noticed that abi1/abi2/hab1/pp2ca mutant plants showed a strong reduction in lateral root growth on 1/2 MS medium ( Figs 8A and S7E ) . The lateral root growth deficiency was also observed on MS medium supplemented with ABA , IAA , or both ( Figs 8A , 8B , S7F and S7G ) ( IAA was used here to accelerate lateral root growth ) . Interestingly , overexpression of GFP-GEF1 partially rescued lateral root length and visible lateral root number defects of abi1/abi2/hab1/pp2ca mutant plants ( Fig 8A and 8B , S7A–S7G Fig ) . Given the reduction of lateral root length ( Fig 8A and 8B ) together with the low GEF1 protein abundance in abi1/abi2/hab1/pp2ca mutant plants ( Fig 5A and 5C ) , we hypothesized that GEF1 together with other GEFs might have the ability to promote lateral root growth . To examine this hypothesis , we studied lateral root growth in wild-type , pyr1/pyl1/2/4 ABA receptor quadruple mutant and GFP-GEF1/WT and GFP-GEF1/pyr1pyl124 overexpression lines . We found that both GFP-GEF1/WT and GFP-GEF1/pyr1pyl124 transgenic lines showed longer average lateral root lengths compared with those in the control wild type ( p = 0 . 002 , two-sample t test , Origin ) and pyr1/pyl1/2/4 mutant lines ( p = 0 . 02 , two-sample t test , Origin ) , respectively ( S8A Fig ) . This result supported that GEF1 has an ability to promote lateral root growth . Interestingly , in the presence of ABA in wild type , the longer lateral root phenotype in GFP-GEF1/WT plants was repressed ( S8B Fig , p = 0 . 26 , two-sample t test , Origin ) . In comparison , lateral root length in GFP-GEF1/pyr1pyl124 was longer than those in pyr1/pyl1/2/4 mutant plants in the presence of ABA ( S8B Fig , p = 0 . 03 , two-sample t test , Origin ) . These data are consistent with a model in which ABA contributes to regulation of RopGEF1-mediated lateral root growth . Furthermore , we also examined lateral root growth in several higher-order gef mutants . Lateral root growth deficiency was not significant in the absence of ABA ( S9C Fig , p = 0 . 1 , two-sample t test , Origin ) but was observed in the gef1/4/10/14 quadruple mutant plants compared with wild-type plants in the presence of ABA ( S9D Fig , p = 0 . 02 , two-sample t test , Origin ) .
Biochemical and structural studies have revealed the molecular mechanism by which ABA receptor proteins that have a higher probability of residing in a monomeric state constitutively bind to and down-regulate PP2Cs [9 , 10 , 60–62] . Eukaryotic receptor signal transduction often relies on regulators that ensure that signal transduction is shut off in the absence of a stimulus [11–13] . An earlier study showed that a small GTPase ROP11 can directly bind to ABI1 and release ABI1 from phosphatase activity inhibition by the monomeric ABA receptor RCAR1/PYL9 [31] . Here , we report that PP2Cs in turn protect RopGEF1 from ABA-mediated degradation . Considering the well-established activation relationship between RopGEFs and ROPs , we propose a RopGEF-ROP-PP2C control loop model that can effectively shut off ABA signal transduction in the absence of ABA ( Fig 8C ) . Under non-stress conditions , basal ABA concentrations in plant cells are low , and monomeric PYL/RCARs bind to PP2Cs in the absence of ABA but with a low affinity [9] . This offers ROPs the opportunity to interfere with the leaky repression of PP2C activity by monomeric PYL/RCARs at low basal ABA concentrations [9 , 10 , 63 , 64] through competitive binding of ROP11 to PP2Cs ( Fig 8C ) . In the absence of ABA , PP2Cs are active and also protect RopGEFs from degradation and thus maintain ROPs in an active form ( Figs 5 and S6 ) . These interactions predict that RopGEF-ROP-PP2Cs form a positive feedback circuit to counteract leaky ABA signal transduction ( Fig 8C ) . Under stress conditions , the ABA content in plant cells increases , and ABA-bound PYL/RCARs bind to PP2Cs with a higher affinity [9 , 10] , which would put ROPs at a competitive disadvantage . Furthermore , ABA signal transduction causes an ABA-mediated rapid degradation of RopGEFs ( Figs 1 , 3 and 5 ) leading to inactivation of ROPs [25 , 26] . As a result , in the presence of ABA , the RopGEF-ROP-PP2Cs circuit collapses rapidly , and PP2Cs are fully inactivated by ABA , releasing SnRK2 protein kinases to trigger ABA responses [1 , 3] . Genetic analyses of higher-order pp2c , gef , and pyl/rcar mutants correlate with this model ( Figs 5–8 and S7–S9 Figs ) . Recent research revealed that ABI1 can interact with the U-box E3 ligases PUB12 and PUB13 and is subsequently degraded through the proteasome . However , PUB12/13 mediated ubiquitination of ABI1 only occurs when ABI1 interacts with PYL/RCAR ABA receptors [65] . This finding may imply that release of ABI1 from constitutive binding to monomeric PYL/RCARs , for example , through ROPs , would be essential to maintain ABI1 activity . The RopGEF-ROP-PP2C circuit loop model exhibits several advantages . First , it could provide a mechanism for dynamic responses to ABA content changes in response to stress [2] and provide a reasonable explanation for binding affinities of PYL/RCARs to PP2Cs [9 , 10] , and for the ability of ROPs to protect PP2C activity [31 , 56] . The proposed control loop may further provide a mechanism for establishing network linkage of ABA signal transduction with other physiological signals and critical cellular processes during plant growth . Physiological signals that affect activities of GEFs and ROPs will impact ABA signal transduction , especially considering that RopGEFs act as a bridge to link various extracellular signals [27 , 28 , 66 , 67] . On the other hand , ABA degradation of GEF1 can also contribute to the response of plants to adverse abiotic stress conditions via down-regulation of plant growth [68 , 69] . ROPs regulate a broad spectrum of cellular processes such as polarized cell growth , cell division , and cell wall restructuring [57 , 59 , 70] . Thus , ABA signal transduction may be able to exert a substantial influence on plant growth through the control of GEF-ROP activities , as shown here for the function of ABA signal transduction in GEF1 regulation of lateral root growth . Based on ABA-mediated particle formation among GEFs , we constructed gef1/4/10 gef1/4/14 triple mutants , and the gef1/4/10/14 quadruple mutant to reduce possible overlapping GEF gene functions . The quadruple mutant gef1/4/10/14 shows enhanced ABA sensitivity in ABA-mediated inhibition of seed germination ( Fig 7F , see also S11 Fig ) . Our findings do not rule out the possibility that additional mechanisms exist to regulate GEF activity by ABA signal transduction , as supported by a recent report that ABA causes degradation of RopGEF2 through the ubiquitin-26S proteasome system [71] . Root branching architecture is a crucial determinant of nutrient and water uptake and lodging resistance of plants . The initiation and postemergence of lateral roots are coordinately regulated by the plant hormones auxin [72 , 73] and ABA [69 , 74] , microRNAs [75] , and environmental responses such as hydrotropism [76] . ABA treatment induces growth quiescence in lateral roots , whereas expression of abi1-1 , which dominantly inhibits ABA signal transduction , leads to a recovery in lateral root growth [69] . In the present study , two GFP-GEF1 overexpression lines , GFP-GEF1/WT and GFP-GEF1/pyr1pyl124 , exhibited longer average lateral root lengths compared to their respective background lines , suggesting that GEF1 has the ability to facilitate lateral root growth ( S8A Fig ) . More importantly , this ability of GFP-GEF1 is inhibited by application of ABA in wild-type but not in pyr1/pyl1/2/4 plants ( S8B Fig ) , implicating a contribution of ABA-mediated down-regulation of GEF1 function . The abi1/abi2/hab1/pp2ca quadruple mutant plants exhibited a severe deficiency in lateral root growth ( Figs 8A , 8B and S7E–S7G ) and low protein abundance of GEF1 ( Fig 5 ) . Furthermore , overexpression of GEF1 in abi1/abi2/hab1/pp2ca plants partially restored lateral root growth ( Fig 8A and 8B ) , which may be attributed to the finding that GEF1 also interacts with additional clade A PP2Cs including HAI1 and AHG1 ( S5B and S5C Fig ) . We speculate that ABA-mediated degradation of GEFs may play an important role in ABA-inhibition of lateral root growth . It was previously shown that AtRAC1/ROP6 activity is rapidly inactivated by ABA in wild-type but not in abi1-1 cells [29] . This could be explained by the present model ( Fig 8C ) , as RopGEFs , the activators of ROPs , are rapidly degraded by ABA ( Figs 1 and 3 ) . GEF degradation leads to the inactivation of ROPs [23–25] . In the abi1-1 mutant , the constitutively active ABI1 phosphatase [4 , 5 , 77 , 78] protects GEF1 from degradation , thus maintaining ROPs in an active form . This is also supported by experimental results showing that both ABA-mediated formation of intracellular GFP-GEF1 particles and degradation of GFP-GEF1 were substantially dampened by overexpression of mCherry-abi1-1 ( S10 Fig ) . Furthermore , our results show that GEF1 directly interacts with ABI1 in Y2H , BiFC , and pull-down assays ( Figs 4 , S4 and S5A ) , and GEF1 undergoes constitutive degradation in abi1/abi2/hab1/pp2ca plants ( Fig 5 ) . Considering the importance of intact PP2Cs for inhibiting GFP-GEF1 degradation ( Fig 5 ) , we speculate that a potential protein kinase can phosphorylate RopGEF1 in response to PP2C inhibition by ABA . RopGEF1 phosphorylation [28] may subject RopGEF1 to degradation ( Fig 8C ) . In the absence of ABA or at low basal ABA concentrations , PP2C-mediated dephosphorylation of RopGEF1 may protect RopGEF1 from degradation , which in turn can prevent leaky ABA signal transduction . Further research will be needed to determine whether the PP2Cs directly dephosphorylate RopGEF1 and to identify protein kinases that may function in signaling of RopGEF1 degradation , and also to determine whether ABA-activated protein kinases such as the SnRK2 protein kinases [79 , 80] are involved in ABA-induced RopGEF1 degradation .
All Arabidopsis thaliana lines were in the Columbia ( Col-0 ) background . The gef1 ( SALK_058164C ) , gef4 ( CS808780 ) , gef10 ( SALK_009456C ) , and gef14 ( CS820474 ) seeds and plasmids CD3-983 , CD3-991 , and CD3-959 were obtained from the Arabidopsis Biological Resource Center . Accession numbers for gef1/4/10 and gef1/4/10/14 mutants are CS69175 and CS69176 . Arabidopsis seeds were surface sterilized in 20% bleach for 30 min followed by four washes with sterile water and sown on 1/2 MS media ( pH 5 . 8 ) supplemented with 1% sucrose and 0 . 8% Phyto Agar . Plates with sterilized seeds were stratified in the dark for 3 d at 4°C and then transferred to the growth room . The growth conditions were as follows: 16/8 light/dark cycle , 80 μmol m-2s-1 light intensity , 22 to 24°C , and 30% relative humidity . 1-wk-old seedlings were transplanted into soil . Coding sequences were amplified from mixed Arabidopsis flower , leaf , and seedling cDNA . Indicated fusion constructs were generated through USER technology [81] . After the sequences were verified , the resulting constructs were transformed into Agrobacterium strain GV3101 . Transgenic lines were generated through the standard floral dip method . Optical density ( OD600 ) of Agrobacterium cells was adjusted to 0 . 3 with buffer ( 10 mM MES , 10 mM MgCl2 , 100 μM acetosyringone , pH 5 . 6 ) . For co-expression , equal volumes of bacterial suspensions were mixed at a final OD600 of 0 . 3 each and infiltrated into 6-wk-old N . benthamiana leaves with a syringe and needle . 48 h after infiltration , the leaves were injected with 50 μM ABA or control buffer ( 0 . 1% v/v EtOH ) and GEF1 fluorescence confocal imaging was conducted 1 h after injection or at the indicated times . GFP-GEF1 and mCherry-γ-TIP constructs were co-expressed in N . benthamiana leaves , and ABA was applied as described above . At 48 h after transfection , mesophyll protoplasts from N . benthamiana leaves were isolated by enzymatic digestion . The enzyme solution contained 1% ( w/v ) cellulase R-10 , 0 . 25% ( w/v ) macerozyme , 10 mM CaCl2 , 20 mM KCl , 400 mM mannitol , and 20 mM MES , pH 5 . 7 . Protoplasts were washed twice by centrifugation at 80 g for 2 min and resuspended in the same solution without enzymes . Vacuoles were released through mixing protoplasts with lysis buffer ( 100 mM malic acid , 3 mM MgCl2 , 0 . 1 mM CaCl2 , pH adjusted to 7 . 5 with bis-Tris propane , and osmolarity adjusted to 500 mOsm with sorbitol ) . Fluorescence signals were detected using an LSM 710 confocal microscope ( Zeiss ) with a 20X objective lens . The following wavelengths were used for fluorescence detection: excitation 488 nm and emission 490–530 nm for GFP; excitation 488 nm and emission 520–550 for YFP; excitation 543 nm and emission 560–620 nm for mCherry and FM4-64; excitation 350 nm and emission 430–480 nm for DAPI . For co-localization analyses , GFP and mCherry fluorescence signals were acquired using the sequential line scanning mode to avoid bleed-through . Confocal imaging data were repeated at least in three independent experiments . For each condition and each experiment , at least five independent cells were analyzed in separate N . benthamiana leaves or Arabidopsis roots when indicated . Representative images are shown in the figures . In order to quantify co-localization results , the linear Pearson ( rp ) and the non-linear Spearman’s rank ( rs ) correlation coefficient ( PSC ) were calculated using FIJI software with an intensity correlation analysis plugin ( http://www . uhnresearch . ca/facilities/wcif/software/Plugins/ICA . html ) . Levels of co-localization for representative areas ( yellow boxes in figures ) are depicted in intensity scatter plots . Calculated PSC values are given in the upper left corner of scatter plots . For time-lapse observations , a rectangular well was made on slides with dow corning high vacuum grease and filled with an ABA-containing 1/2 MS solution . Roots of seedlings or N . benthamiana leaf discs were mounted in solution with a coverslip for confocal microscopy . For Arabidopsis leaves , young 2-wk-old plants were used and the abaxial epidermis was imaged in intact leaves . Single confocal focal planes and confocal images were recorded at the indicated times . Y2H assays were performed with the USER-modified pGBT9 and pGADGH vectors ( Clontech ) . Indicated bait and prey constructs were transformed into Yeast PJ69-4A cells and selected on SD-L-W medium . Yeast colonies were re-streaked onto new SD-L-W plates and incubated 1 or 2 d . Successfully transformed clones were incubated in SD-L-W liquid medium overnight and then the OD600 of cultures was adjusted to 1 with sterile water . A series of 2 μl 10-fold dilutions of transformants were spotted on SD-L-W and SD-L-W-H supplemented with 3 mM 3-amino-1 , 2 , 4-triazole ( 3-AT ) , and grown for 6 d . The coding sequence of RopGEF1 was cloned into the pGEX6P-1 ( GE Healthcare ) to generate a GST-GEF1 fusion construct , and coding sequences of ABI1 , OST1 , and ROP11 were cloned into the modified pET52strep-II vector to generate Strep-II tag labeled ABI1 , OST1 , and ROP11 fusion constructs . After sequences were verified , the constructs were transformed into E . coli Rosetta ( DE3 ) pLysS ( Novagen ) cells , and expression of the fusion protein was induced by 0 . 5 mM IPTG ( isopropylthio-β-galactoside ) at 18°C overnight ( 1 L overnight bacterial cultures were used for purification of StrepII-ABI1 ) . GST and strep-II fusion proteins were purified using Gluotathione SepharoseTM fast flow ( GE healthcare , Pittsburgh , US ) and Strep-Tactin resin ( IBA , Goettingen , Germany ) , respectively , according to the manufacturers’ instructions . Pull-down assays were performed according to [82] . Bound proteins were eluted , fractionated by 10% SDS-PAGE , and subjected to immune-blot analysis using Strep-Tactin HRP conjugate ( IBA , Göttingen , Germany ) . For western blot assays , 10-d-old Arabidopsis seedlings grown on 1/2 MS medium were transferred to 1/2 MS liquid media for 1 h followed by hormone or chemical treatments for the indicated times . Seedlings were ground in liquid nitrogen and then resuspended in an extraction buffer containing 25 mM Tris-HCl ( pH 7 . 5 ) , 150 mM NaCl , 10% glycerol , 10 mM DTT , 1 mM EDTA , 0 . 5% Triton X-100 , 1 mM PMSF , 10 mM NaF , and 1 × protease inhibitor cocktail ( Roche ) and incubated on ice for 30 min . Cell debris was pelleted by centrifugation at 15000 g for 20 min at 4°C; the supernatant was transferred to new tubes . Protein concentration was determined using the Bradford protein assay kit ( IBI Scientific , Peosta , US ) . 20 μg proteins were separated on 10% SDS-PAGE gel , blotted onto a nitrocellulose membrane ( Milli-pore ) , probed with anti-GFP antibody ( Life Technology ) overnight at 4°C , and then incubated with a goat anti-rabbit HRP-conjugated secondary antibody ( Bio-Rad ) for 1 h at room temperature . Membranes were incubated with chemiluminescence reaction solution ( Pierce , Rockford , US ) and western blot signal was detected using a typhoon Imager FLA 7000 ( GE healthcare , Pittsburgh , US ) . Agrobacterium cells containing RK19 ( to reduce the silencing of transgenes ) were mixed with those containing pUBQ-Flag-GFP-GEF1 or pUBQ-myc-mCherry-PP2C constructs at a ratio of 1:1 . Subsequently , the mixture was co-injected into N . benthamiana leaves . At 48 h after infiltration , N . benthamiana leaves were ground in liquid nitrogen and homogenized in protein extraction buffer ( 25 mM Tris-HCl , pH 7 . 5 , 150 mM NaCl , 10% glycerol , 0 . 1% Nonidet P-40 , 1 mM EDTA , 10 mM DTT , 1 mM phenylmethylsulfonyl fluoride , and 1 × complete cocktail of protease inhibitors , 2 ml buffer/g leaves ) . Lysates were vigorously votexed for 30 s , then incubated in ice for 30 min . After centrifugation at maximal speed 15000 g for 15 min at 4°C , the supernatant was transferred into a new tube and passed through a 0 . 45 μm filter ( Millipore ) . 30 μl of flag magnetic beads ( Sigma ) were added to 5 mg of total proteins , and the mixture was incubated for 4 h at 4°C . The precipitated samples were washed four times with the protein extraction buffer ( rotated at 4° C for 5 min , beads were collected with DynaMagTM-2 [Life Technology] ) and then eluted by boiling in 2 × SDS loading buffer for 5 min . Immunoprecipitation products were detected by immunoblotting with rabbit monoclonal flag ( Sigma ) or mouse monoclonal myc antibody ( Life Technology ) . Fifteen 10-d-old seedlings were collected into 2 ml tubes with steel beads , frozen in liquid nitrogen , and ground with a Mixer Mill MM400 ( Retsch ) . Total RNA was extracted using the SpectrumTM Plant Total RNA kit ( Sigma ) and quantified . Approximately 3 μg RNA samples were treated with 1 μl DNase I ( NEB ) for 30 min and converted to cDNA using a First-Strand cDNA Synthesis kit ( GE Healthcare ) . Synthesized cDNA was diluted four times and 2 μl was used as PCR templates . qPCR analyses were performed on a plate-based BioRad CFX96 qPCR System using SYBR Select Master Mix for CFX ( Applied Biosystems ) with gene-specific primers . GEF1fwd: tgcttgccgaaatggagattccc; GEF1rev: agacattccttcccgctcttgg; GAPCfwd: tcagactcgagaaagctgctac; GAPCrev: cgaagtcagttgagacaacatcatc . After surface sterilization of the seeds , 70 seeds of each genotype were sowed on 1/2 MS plates supplemented with ABA . Stratification was conducted in the dark at 4°C for 3 d . Radical emergence was recorded at the indicated times . At 6 d , seedlings with expanded green cotyledons were scored as the percentage of seeds . For root elongation assays , seeds were stratified for 3 d in the dark at 4°C ( for experiments including abi1abi2hab1pp2ca quadruple mutant , all seeds were stratified for 10 d ) , sown , and grown on vertically oriented 1/2 MS plates for 4 d . Twenty seedlings with similar primary root length were transferred onto new 1/2 MS plates lacking or supplemented with the indicated concentrations of ABA . The plates were scanned after 6 d of growth , and primary root lengths were measured with ImageJ ( http://imagej . nih . gov/ij/download . html ) . Twenty 4-d-old seedlings grown on 1/2 MS medium with similar primary root lengths were transferred onto new 1/2 MS plates lacking or supplemented with the indicated concentrations of ABA and IAA . The plates were scanned after the indicated additional times of growth , and primary root lengths and total lateral root lengths ( lateral roots longer than 0 . 3 mm were measured and counted ) were measured with ImageJ . Average lateral root length was defined as total lateral root length divided by lateral root number for each root . Lateral root ( LR ) number/centimeter are lateral root number divided by primary root length . Because different hormones have different effects on lateral root growth , it is difficult to compare the effect of different hormone treatments at the same time . For example , in the present study , lateral roots were too dense to be distinguished on auxin or auxin plus ABA medium plates when lateral roots were imaged >4 d after transfer of seedlings , and the visible lateral roots on MS medium have just emerged . Therefore , plates were scanned and lateral roots were measured at the indicated times when lateral root lengths could be accurately measured . Comparisons were made among genotypes within each group with identical hormone treatments . All relevant data are within the paper and its Supporting Information files .
|
The plant hormone abscisic acid ( ABA ) is critical to plant development and stress responses . The ABA signaling cascade is comprised of ABA receptors , negatively regulating PP2C protein phosphatases and transducing protein kinases . Biochemical assays have indicated that certain ABA receptors constitutively bind to and inhibit PP2C protein phosphatases even without ABA . This finding has suggested that leaky receptor signal transduction in the absence of ABA could occur . Small GTPases named ROPs are negative regulators of ABA signal transduction and maintain PP2C protein phosphatase activity . However , whether and how the ABA signal can remove the inhibition of ABA signaling by ROPs remains elusive . The GTP exchange factor , RopGEF1 , is an activator of ROP small GTPases . We show that the subcellular localization of RopGEF1 rapidly changes in response to ABA . RopGEF1 is sequestered via the endosome-prevacuolar compartment pathway and is degraded . Interestingly , we have found that ABI1 , a PP2C protein phosphatase , directly interacts with RopGEF1 . Moreover , PP2Cs , which are active in the absence of ABA , protect the protein stability of RopGEF1 . These findings point to a RopGEF1-ROP-ABI1 control loop model that could protect against leaky receptor signaling , and which facilitates ABA signal transduction when ABA is produced in response to stress .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"plant",
"anatomy",
"plant",
"growth",
"and",
"development",
"enzymes",
"lateral",
"roots",
"enzymology",
"phosphatases",
"developmental",
"biology",
"guanine",
"nucleotide",
"exchange",
"factors",
"plant",
"science",
"seedlings",
"plants",
"cellular",
"structures",
"and",
"organelles",
"proteins",
"subcellular",
"localization",
"biochemistry",
"signal",
"transduction",
"plant",
"biochemistry",
"plant",
"roots",
"cell",
"biology",
"root",
"growth",
"biology",
"and",
"life",
"sciences",
"cell",
"signaling",
"organisms",
"signaling",
"molecules"
] |
2016
|
Release of GTP Exchange Factor Mediated Down-Regulation of Abscisic Acid Signal Transduction through ABA-Induced Rapid Degradation of RopGEFs
|
BK polyomavirus ( BKV or BKPyV ) associated nephropathy affects up to 10% of kidney transplant recipients ( KTRs ) . BKV isolates are categorized into four genotypes . It is currently unclear whether the four genotypes are also serotypes . To address this issue , we developed high-throughput serological assays based on antibody-mediated neutralization of BKV genotype I and IV reporter vectors ( pseudoviruses ) . Neutralization-based testing of sera from mice immunized with BKV-I or BKV-IV virus-like particles ( VLPs ) or sera from naturally infected human subjects revealed that BKV-I specific serum antibodies are poorly neutralizing against BKV-IV and vice versa . The fact that BKV-I and BKV-IV are distinct serotypes was less evident in traditional VLP-based ELISAs . BKV-I and BKV-IV neutralization assays were used to examine BKV type-specific neutralizing antibody responses in KTRs at various time points after transplantation . At study entry , sera from 5% and 49% of KTRs showed no detectable neutralizing activity for BKV-I or BKV-IV neutralization , respectively . By one year after transplantation , all KTRs were neutralization seropositive for BKV-I , and 43% of the initially BKV-IV seronegative subjects showed evidence of acute seroconversion for BKV-IV neutralization . The results suggest a model in which BKV-IV-specific seroconversion reflects a de novo BKV-IV infection in KTRs who initially lack protective antibody responses capable of neutralizing genotype IV BKVs . If this model is correct , it suggests that pre-vaccinating prospective KTRs with a multivalent VLP-based vaccine against all BKV serotypes , or administration of BKV-neutralizing antibodies , might offer protection against graft loss or dysfunction due to BKV associated nephropathy .
The process of kidney transplantation has been revolutionized since the first successful case in identical twins more than 5 decades ago [1] , [2] . Since then , the use of immunosuppressants such as cyclosporine has made renal allografts a viable clinical option [3] , but the process still has many challenges , including the management of chronic and acute immune-mediated rejection of the allograft , nephrotoxicity from immunosuppressants and antiviral drugs , and controlling opportunistic infections . To balance these factors , clinical guidelines for the treatment of kidney transplant recipients ( KTRs ) generally suggest the use of intensive immunosuppression during the initial stages of the process , followed by a diminished dose of immunosuppressants if there are no signs of acute rejection by 2–4 months after transplantation [4] . In addition to the potential problem of immunological rejection of the allograft , between 1 and 10% of KTRs develop nephropathy associated with a non-enveloped DNA virus species called BK polyomavirus ( BKV or BKPyV ) [5]–[8] . Serological and PCR-based studies indicate that nearly all human beings are chronically infected with BKV [9] , [10] . Although chronic BKV infection of the urinary tract is not known to be associated with overt clinical symptoms in healthy individuals , opportunistic replication of the virus in KTRs can lead to graft dysfunction or loss [11] . BKV can also cause a bladder condition known as hemorrhagic cystitis in bone marrow transplant recipients and in cancer patients treated with the immunosuppressant cyclophosphamide [12] , [13] . Clinical guidelines for these conditions therefore recommend regular monitoring of serum or urinary BKV viral load and reduction of immunosuppression if signs of uncontrolled BKV replication are observed [4] . In pediatric KTRs , being BKV seronegative prior to transplantation correlates with the risk of developing BKV associated nephropathy ( BKVN ) [14] . In adult KTRs , at least one study suggested a significant correlation between donor BKV seroreactivity and the risk of urinary shedding of BKV in KTRs [15] . However , BKV seroprevalence is high in most adult populations and a variety of other studies have not uncovered clear correlations between KTR seroresponsiveness to BKV and resistance to BKV viremia or BKVN [16]–[18] and reviewed in [19] . Monitoring of pre-transplant BKV serology is not usually performed , based largely on the idea that it is not an effective indicator of susceptibility to BKVN . This logic has also been used to infer that vaccine-induced boosting of BKV-specific antibody responses in prospective KTRs would be unlikely to confer protection against BKVN . BKV isolates can be grouped into four genetically distinct subspecies ( genotypes ) [20]–[22] . While chronic infection with BKV genotype I ( BKV-I ) is believed to be common almost to the point of ubiquity in all human populations worldwide , PCR-based studies suggest that BKV genotypes II , III , and IV only infect a minority of adults . The incidence of BKV-IV infection varies among different populations , with estimates ranging from <5% of BKV isolates detectable in urine specimens contributed by Sub-Saharan African subjects to 54% of detected isolates from Northeast Asian subjects [23] . The prevalence of BKV genotypes II and III appears to be lower than BKV-IV [9] . However , a complicating factor in these studies is that many commonly-used PCR primer pairs detect BKV genotype I more efficiently than other genotypes [24] . The extent of serological cross-reactivity among BKV genotypes is unclear . Prior serological studies of KTRs have monitored BKV-specific serum antibody responses using ELISAs where the target antigens are recombinant virus-like particles ( VLPs ) assembled from the BKV major capsid protein VP1 . A recent study [25] using VLPs based on a BKV-III isolate reported serological findings similar to earlier studies using BKV-I VLPs [19] . Furthermore , one VLP-based study found a significant correlation between BKV-II and BKV-III seroresponsiveness in human subjects [26] . Although past serological studies might seem to suggest a lack of distinct BKV serotypes , a limitation of VLP ELISAs is that they simultaneously detect antibodies that can neutralize BKV infectivity and non-neutralizing antibodies . Viruses are thought to be under greater selective pressure to accumulate evasive mutations in capsid epitopes recognized by antibodies that can neutralize infectivity . In contrast , there is presumably less selective pressure for viruses to accumulate mutations in virion surfaces recognized by antibodies that don't neutralize infection . Consequently , epitopes bound by non-neutralizing antibodies are more likely to be conserved among related viral genotypes , while neutralizable epitopes are more likely to be divergent . This poses a problem for ELISA methodology , since measurement of non-neutralizing antibody binding may obscure the existence of a subset of serotype-specific neutralizing antibodies . Consistent with this theory , we have previously shown that serological assays that monitor antibody-mediated neutralization of a different family of non-enveloped viruses , the Papillomaviridae , offer a more effective way to discriminate among human papillomavirus ( HPV ) serotypes than VLP-based ELISAs [27] . Neutralization-based serological methods also appear to offer an accurate reflection of the clinical efficacy of recently developed VLP-based vaccines against HPV [28] , [29] . In a 1989 study using various cell culture-adapted BKV isolates , Knowles and colleagues demonstrated that serum antibodies from animals immunized with virions of one BKV genotype are less effective for in vitro neutralization of BKV isolates from the other three genotypes [22] . Whether this serological difference also holds true for people with chronic BKV infections is not known . For example , some people with HIV-1 infections eventually develop antibody responses capable of blocking the infectivity of a wide range of HIV-1 genotypes . Such broadly cross-neutralizing antibody responses generally take many years to develop and are not typically observed in acutely immunized animals ( reviewed in [30] and [31] ) . To investigate whether BKV genotypes can be divided into distinct neutralization serotypes in naturally infected humans , we developed high-throughput serological assays to monitor antibody-mediated neutralization of the infectivity of BKV-I and BKV-IV reporter vectors ( pseudovirions ) . Our results show that BKV-I and BKV-IV are distinct serotypes with respect to functionally neutralizing serum antibodies . Using neutralization assay methodology , we monitored the development of BKV-I and BKV-IV seroresponsiveness in a cohort of KTRs . The data show that a substantial fraction of KTRs who lacked detectable BKV-IV neutralizing antibody responses one week after transplantation experienced acute BKV-IV-specific seroconversion during the first year after transplantation . This may reflect a de novo BKV-IV infection arising from the engrafted kidney . The findings raise the possibility that VLP-based vaccination of candidate organ transplant recipients who are initially naïve against specific BKV serotypes might confer protection against pathological forms of BKV replication that are sometimes associated with the implementation of immunosuppressive therapy .
Virus-like particles ( VLPs ) , including BKV VLPs [32] , can be potently immunogenic when administered as vaccines in animal model systems ( reviewed in [33] ) . To generate BKV VLP vaccine immunogens , we co-expressed the VP1 , VP2 and VP3 capsid proteins of the BKV-I isolate KOM-5 or the BKV-IV isolate A-66H via transfection of the human embryonic kidney-derived cell line 293TT [34]–[37] . The resulting VLPs were purified by ultracentrifugation through Optiprep gradients . Five µg of purified BKV-I or BKV-IV VLPs were mixed with complete Freund's adjuvant and administered separately to sets of six mice . Serum antibody responses were assayed four weeks after a single vaccination using ELISA plates coated separately with BKV-I or BKV-IV VLPs . All the mice showed high titer serum antibody responses in ELISAs against the cognate BKV VLP type , except for one BKV-I vaccinated mouse which appeared to be relatively non-responsive ( Figure 1 , top panel ) . The sera exhibited varying amounts of cross-reactivity against the non-cognate BKV . The average ratio of cognate BKV to non-cognate BKV titer was 21 for mice immunized with BKV-I and 110 for mice immunized with BKV-IV ( Figure 1 , bottom panel ) . It has recently become possible to generate reporter vectors ( also known as pseudoviruses ) based on BKVs [35] , [36] . These recombinant production systems made it possible for us to generate infectious capsids composed of the VP1/2/3 capsid proteins of BKV primary isolates of genotypes I and IV that are not otherwise culturable . Using reporter vector-based assays , we titered the neutralizing potency of sera from BKV-I or BKV-IV vaccinated mice . As expected , the neutralization assays showed a significantly greater degree of BKV type-specificity compared to the ELISAs . The median cognate versus non-cognate neutralizing titer ratio was 910 for mice immunized with BKV-I and 620 for mice immunized with BKV-IV ( Figure 1 ) . A comparison of ELISA values to neutralization assay values is shown in Figure S1 . To test the possibility that a booster vaccination might alter the degree of cross-neutralization of the two BKV types , we administered the mice a second dose of cognate VLPs in incomplete Freund's adjuvant one month after priming . We then performed repeat serology a total of two months after the initial priming dose . Hyperimmune sera from the boosted animals showed neutralizing titer ratios similar to the initial testing ( data not shown ) , suggesting that boosting did not have a major effect on cross-neutralization profiles . Sera from 48 healthy adults with a median age of 52 . 5 years were assessed for reactivity to BKV-I and BKV-IV in ELISAs . Eighty-three percent of the volunteers were seropositive in the BKV-I ELISA ( Figure 2 , top panel ) , a prevalence similar to what has been reported in the literature [9] , [10] . Sixty-five percent of volunteers scored seropositive in the BKV-IV ELISA . This is in contrast to the purportedly much lower prevalence of BKV-IV infection , but is consistent with the possibility that the BKV-IV ELISA detects cross-reactive antibodies elicited by BKV-I infection ( and perhaps vice versa ) [31] . The significant correlation between individual subjects' BKV-I and BKV-IV ELISA titers ( Spearman r = 0 . 69 , p<0 . 0001 ) is also consistent with the possibility that the BKV-I and BKV-IV ELISAs exhibit a significant degree of cross-reactivity when used for analysis of human sera . We next applied the BKV-I and BKV-IV neutralization assays to the human serum samples . Only 3 volunteers ( 6% ) were negative for BKV-I neutralization , while 37 ( 77% ) volunteers scored seronegative in the BKV-IV neutralization assay ( Figure 2 , bottom panel ) . The increased rate of seropositivity for BKV-I compared to ELISA suggests that the neutralization assay is more sensitive . In contrast to the ELISA results , there was not a statistically significant correlation between the subjects' BKV-I and BKV-IV neutralizing titers . There were two individuals with BKV-I neutralizing titers of >100 , 000 whose sera did not detectably neutralize BKV-IV at the lowest tested dilution ( 1∶100 ) . This indicates that these individuals displayed BKV type-specificity ratios of at least 1 , 000 . Overall , the results for the human sera appear to confirm the observations using murine sera , showing that neutralization assays offer a greater degree of sensitivity and specificity for serological analysis of exposure to BKV-I and BKV-IV . To gain insight into which VP1 amino acids might dictate BKV neutralization serotypes , we aligned full-length non-identical VP1 peptide sequences available via GenBank ( Figure S2 ) . With respect to the BKV-I consensus , BKV-IV isolates tend to carry a variety of substitutions: E61N , N62D , F66Y , K69R , S71T , N74T , D75A , S77D , E82D , Q117K , H139N , I178V , F225Y , A284P , R340Q , K353R , and L362V . Mapping of these BKV-I/BKV-IV variant residues onto homologous positions in the X-ray crystal structures of JCV [38] and SV40 [39] suggests that , with the exception of positions 117 , 225 , 284 , and 340 , each of these BKV-I/BKV-IV variant residues is likely to be exposed on the exterior surface of the capsid . With the exception of residues 353 and 362 , which are exposed along the floor of the canyons between capsomer knobs , all the exposed variations map to sites on the apical surface and apical rim of the capsomer knob . Many of the variations are adjacent to residues predicted to be involved in binding the cellular glycolipids that serve as receptors during BKV infectious entry [40] , [41] . This is consistent with the idea that BKV-I/BKV-IV variations may alter epitopes recognized by antibodies that neutralize infectivity via steric occlusion of the receptor binding site . In addition to the differences between BKV-I and BKV-IV , we noted several positions that differ stereotypically among BKV-I subtypes . For example , BKV subtype Ib-2 isolates tend to carry V42L , E82D , D175E , V210I , R340K , and L362V differences , with respect to subtypes Ia and Ib-1 . Likewise , subtype Ic isolates frequently carry E20D , F225L , and R340K differences . Although these intra-genotype-I surface variations are chemically subtle , it is conceivable that the differences reflect selective pressure to escape neutralizing antibodies . An important goal of future studies will be investigation of the possibility that BKV genotype I encompasses more than one neutralization serotype . It will also be interesting to learn whether the reportedly less prevalent BKV genotypes II and III are serologically distinct from one another and/or distinct from genotypes I and IV . Sera collected from 108 KTRs at time points of roughly 1 , 4 , 12 , 26 and 52 weeks post-transplantation were tested using the BKV-I and BKV-IV neutralization assays . Testing of more than 500 samples at a full set of 10 serial dilutions would be expensive and logistically challenging . Therefore each serum sample was tested at 4 dilutions: 100 , 500 , 5 , 000 , and 50 , 000 . Because of this lack of full serial dilution , we elected to use a more stringent 95% neutralization cutoff of for individual data points . Neutralization assay results for individual subjects are shown in Figures 3 and S3 . At study entry only 5 ( 5% ) patients scored seronegative ( i . e . , <95% neutralizing at the 1∶100 serum dilution ) in the BKV-I neutralization assay ( Table 1 ) . In contrast , there were 53 ( 49% ) initially BKV-IV seronegative patients . In an initial analysis , seroconversion was defined as a change from seronegative at study entry to at least 95% neutralization at the 1∶500 serum dilution at any subsequent time point . By this standard , all 5 ( 100% ) of the initially BKV-I seronegative patients seroconverted for BKV-I , while 23 ( 43% ) of the initially BKV-IV seronegative patients seroconverted for BKV-IV ( Table 1 ) . The average BKV type-specificity ratio for sera from immunized mice was 1 , 359 ( Figure 1 ) . Two human subjects likewise showed type-specificity ratios >1000 ( Figure 2 ) . To address the possibility that BKV-IV neutralization might be partly attributable to cross-reactivity of high titer antibody responses elicited by BKV-I , we applied a more stringent definition of seroconversion , in which the ratio of the BKV-I titer versus the BKV-IV titer must be 1 , 000 or less to be considered a clear type-specific seroconversion event . Using these stricter criteria , 12 ( 23% ) of the initially BKV-IV negative patients seroconverted within a year of transplantation ( Table 1 ) . Based on the results shown in Figures 1 and 2 , the occurrence of BKV type-specificity ratios of 10 or less seems highly unlikely . Five patients ( 5% ) underwent BKV-IV-specific seroconversion by the extremely strict criterion of having a BKV-I to BKV-IV titer ratio ≤10 . A previous study of this set of subjects used nested PCR followed by restriction fragment analysis to determine which BKV genotype was shed in the urine or blood of a subset of patients who became viruric during the course of the study [15] . Two of four patients previously found to have shed BKV-IV DNA in their urine during or after the onset of viruria seroconverted for BKV-IV neutralization by the extremely strict definition of seroconversion ( Figures 3 and S3 ) . Interpretation of the BKV genotyping data is restricted by two caveats . First , the nested PCR method used for the genotyping is likely to detect only the most abundant BKV genotype present in the sample . This problem has recently been highlighted by Luo and colleagues in a report showing that BKV-IV and chimeric quasispecies containing BKV-IV-related sequences are often present as minority sequences in urine samples from KTRs and healthy subjects with viruria [42] . An additional problem for BKV genotype analysis is that virions found in blood or urine may originate from the recipient's bladder epithelium [43]–[45] or original kidneys . Such virions might not reflect the infection status of the engrafted kidney . Interestingly , Randhawa and colleagues have reported a higher prevalence of BKV genotype IV ( 5/25 ( 20% ) ) , as well as ambiguous BKV genotypes ( 5/25 ( 20% ) ) in biopsy material from engrafted kidneys affected by interstitial nephritis [46] . Establishing clear relationships ( or lack thereof ) between BKV type-specific serological titer and the presence of various BKV genotypes in patients at risk of BKVN will likely require deep sequencing of BKV DNA amplified using primers that target conserved portions of the BKV genome [24] . Ideally , such analyses would include biopsy specimens from engrafted kidneys . On average , the patients' BKV-I and BKV-IV neutralizing titers both increased substantially by one year after renal transplantation ( Figure 4 ) . In some instances , titer increases occurred even in patients who showed moderate neutralizing antibody titers at study entry ( Figures 3 and S3 ) . This result mirrors data that Bohl and colleagues obtained for this same set of sera using BKV-I VLP ELISAs [16] . It is unclear whether these titer increases reflect new cycles of infectious cell-to-cell spread or emergence of virion production from a latent reservoir , perhaps arising from the recipient's bladder epithelium or original kidneys .
In this report , we show that BKV-I and BKV-IV are distinct serotypes with respect to neutralizing antibody responses in human subjects . Until now , this concept would have been obscured by that fact that traditional VLP-based ELISAs , which have been used for prior investigations of BKV serology in human subjects , do not offer an accurate measurement of BKV serotype-specific neutralizing antibody responses . Our longitudinal neutralization-based analysis of archived sera from 108 KTRs indicates that roughly half of the subjects were initially BKV-IV naïve at the time of transplantation . Roughly half of the initially seronegative subjects went on to show evidence of BKV-IV-specific seroconversion during the first year after transplantation . A simple model for this result would be that half of all adults are latently infected with BKV-IV and transplantation of a latently BKV-IV-infected kidney leads to opportunistic viral replication in BKV-IV-naïve KTRs . The concept that BKVN is a consequence of a de novo BKV infection arising from the engrafted kidney could help explain why the condition rarely affects recipients of other organ types , despite the use of similar immunosuppressive regimens [15] . An important implication of our findings is the possibility that induction of an effective BKV-IV-neutralizing antibody response prior to transplantation might protect some KTRs against outgrowth of BKV-IV harbored in the engrafted kidney . This might , in turn , prevent the development of pathological forms of BKV replication . The idea that BKV-IV may be disproportionately involved in BKVN is supported by studies showing that BKV viremia and viruria in KTRs is frequently attributable to BKV-IV [47]–[50] . Our results show that a single adjuvanted dose of BKV-IV VLPs can induce high titer BKV-IV-neutralizing antibody responses in experimentally vaccinated mice . This suggests that a BKV-IV VLP vaccine would likely be an effective way to elicit neutralizing antibody responses in prospective KTRs prior to the implementation of immunosuppressive therapy . The potential value of pre-vaccination of KTRs could also extend to instances where donors and recipients are discordant for other BKV serotypes . In theory , elicitation of a broadly neutralizing antibody response against all BKVs might be accomplished using a vaccine containing VLPs based on multiple BKV serotypes . A precedent for this idea can be found in the vaccines Cervarix and Gardasil , which elicit neutralizing antibody responses against multiple HPV serotypes using mixtures of VLPs ( reviewed in [51] ) . The clinical efficacy of current HPV vaccines correlates strongly with serological measurements using HPV neutralization assays [29] . This suggests that BKV neutralization assays , such as those reported here , could likewise serve as a useful proxy for the efficacy of candidate BKV vaccines . Several previous studies have employed intravenous infusions of purified immunoglobulins ( IVIG ) in an attempt to suppress humoral responses to the engrafted organ and to possibly offer antibody-based suppression of BKV replication . The results of IVIG studies have been mixed [52]–[54] . In theory , this may have been due to differing levels and serotype specificities of BKV-neutralizing antibodies in various immunoglobulin preparations . A possible alternative approach to polyclonal IVIG would be to administer a cocktail of humanized monoclonal antibodies ( mAbs ) capable of neutralizing all BKV serotypes . Neutralization-based serology approaches should be useful for future determination of the total number of BKV serotypes and for the development of candidate anti-BKV therapeutic antibodies .
Samples from 108 renal transplant subjects from the “Randomized Prospective Controlled Clinical and Pharmacoeconomic Study of Cyclosporine vs . Tacrolimus in Adult Renal Transplant Recipients” of the Washington University in St . Louis School of Medicine were used . For the original study , The Human Studies Committee of the Washington University School of Medicine approved the protocol and informed consent was obtained from all participants . For this study , the samples were assigned random identifier symbols prior to analysis at the National Cancer Institute ( NCI ) . All animal work was approved by the Animal Care and Use Committee of the NCI , according to the guidelines of the Association for Assessment and Accreditation of Laboratory Animal Care International . Procedures were carried out in accordance with the eighth edition of the National Research Council of the National Academies' Guide for the Care and Use of Laboratory Animals . All efforts were made to minimize animal suffering . Eight week old female BALB/cAnNCr mice were immunized once subcutaneously with 5 µg of BKV-I or BKV-IV virus-like particles ( VLPs , see below ) emulsified in complete Freund's adjuvant ( CFA , Sigma ) . Sera were collected four weeks after immunization . The animals were kept under specific pathogen-free conditions in compliance with institutional guidelines at the National Cancer Institute ( NCI ) . The clinical protocols used in this study have previously been described in detail [15] , [17] , [55] . Briefly , patients were given an immunosuppressive regimen , which was reduced if viremia was detected . Serum samples were collected at roughly 1 , 4 , 12 , 26 and 52 weeks post-transplantation . None of the patients were observed to suffer from BKVN during the course of the collection period . A previously-described set of sera from healthy subjects visiting U . S . plasma donation centers were purchased from Equitech Bio and Innovative Research [56] . BKV reporter vectors ( pseudovirions ) were generated as previously described [56] . BKV-I reporter vectors were produced using plasmid pCAG-BKV [35] , which encodes the capsid proteins of BKV isolate KOM-5 . KOM-5 is classified as a BKV type I subtype b-1 ( Ib-1 ) genotype [57] . For BKV-IV particles , the sequence of BKV isolate A-66H ( subtype IVc-2 ) [58] was used to design synthetic codon-modified VP1 , VP2 and VP3 genes [36] . Capsid protein expression plasmids were co-transfected with a reporter plasmid encoding Gaussia luciferase ( phGluc ) into 293TT cells [34] . Forty eight hours after transfection , the cells were suspended at >100 million cells/ml in PBS and lysed by addition of 0 . 5% Triton X-100 , and RNase A/T1 cocktail ( Ambion ) . The lysate was incubated at 37°C overnight to allow capsid maturation , then clarified at 5 , 000× g . Reporter vector particles were purified out of the clarified supernatant by ultracentrifugation through a 27–33–39% iodixanol ( Optiprep , Sigma ) step gradient [59] . For generation of VLPs , 293TT cells were transfected with VP1/2/3 expression plasmids without any reporter plasmid . Two days after transfection , the cells were lysed with 0 . 5% Triton X-100 in DPBS supplemented with 25 mM ammonium sulfate , Benzonase ( Sigma ) , Plasmid Safe ( Epicentre ) and 1 . 2 U/ml neuraminidase V ( Sigma #N2876 ) [60] , [61] . The lysates were incubated at 37°C overnight , then adjusted to 0 . 85 M NaCl , clarified as above and subjected to purification over Optiprep gradients . Maps of the plasmids used in this work and detailed protocols for reporter vector and VLP production can be found at the following website: <http://home . ccr . cancer . gov/LCO/> . Plasmids are available through <http://www . AddGene . org/> . Immulon H2B plates ( Thermo ) were coated with 15 ng/well of VLPs in PBS overnight . PBS with 1% non-fat dry milk ( blotto ) was used to block the coated plates for 2 hours at room temperature , with orbital rotation . Sera from mice and healthy human subjects were serially diluted in blotto and incubated on blocked plates at room temperature for 1 hour , with orbital rotation . Washing was performed with PBS . Horseradish peroxidase conjugated goat anti-mouse IgG ( BioRad ) or donkey anti-human IgG ( Jackson ) secondary antibody diluted 1∶7500 in blotto was used to detect bound antibodies . The plates were incubated with ABTS substrate ( Roche ) and absorbance was read at 405 nm with a reference read at 490 nm . The effective concentration 10% ( EC10 ) was calculated using Prism software ( GraphPad ) to fit a curve to the OD values for each serially diluted serum sample . The top of each dose-response curve was constrained based on the average of the calculated plateau maximum ( Bmax ) values for strongly reactive sera . The Bmax value was typically an OD of around 2 . 0 , such that the EC10 value can be considered comparable to an OD cutoff value of 0 . 2 . Neutralization assays were performed using previously reported methods [56] . Briefly , 293TT cells were seeded at a density of 3×104 per well and allowed to attach for 3–5 hours . Sera from mice and human subjects were serially diluted , and sera from renal transplant patients were tested in separate assays at 4 different dilutions: 1∶100 , 1∶500 , 1∶5 , 000 and 1∶50 , 000 . Dilutions were performed in cell culture medium ( DMEM without phenol red supplemented with 25 mM HEPES , 10% heat-inactivated fetal bovine serum , 1% MEM non-essential amino acids , 1% Glutamax and 1% antibiotic-antimycotic , all from Invitrogen ) . Twenty-four µl of diluted sera were mixed with 96 µl of diluted reporter vector stock and placed at 4°C for 1 hour . One hundred µl of this mixture were added to the cells for 72 hours . Conditioned supernatants ( 25 µl ) were harvested into white 96-well luminometry plates ( Perkin Elmer ) . Fifty µl of Gaussia Luciferase Assay Kit substrate ( NEB ) were injected immediately prior to luminometry using a BMG Labtech Polarstar Optima luminometer . For sera from mice or from healthy human subjects , 50% neutralizing titers ( EC50 ) were calculated based on dose-response curves with top and bottom values constrained to the average values of “no serum” and “no reporter vector” control wells , respectively . For transplant patients , the following criteria for seropositivity and seronegativity were adopted: sera were considered negative at entry if the 1∶100 dilution did not mediate at least a 95% reduction in Gaussia luciferase activity ( measured in relative light units , RLUs ) relative to the no serum control condition ( i . e . , >95% neutralization of the reporter vector ) . Seroconversion refers to subjects who scored seronegative at study entry but whose sera became >95% neutralizing at the 1∶500 dilution at any subsequent time point . A stricter definition of seroconversion , accounting for the possible low-level cross-type neutralization , added the stipulation that the 95% neutralizing titer for BKV-IV differed from the BKV-I neutralizing titer by less than 1 , 000-fold . VP1 protein sequences from BKV strains indicated in Figure S2 were downloaded from GenBank . ClustalW alignments were performed with MacVector software version 11 . 1 . 2 using a Gonnet series matrix . Structural modeling of BKV VP1 amino acid variations was performed by aligning the sequences of JCV or SV40 VP1 to BKV , followed by inspection of homologous positions of interest in the JCV or SV40 VP1 X-ray crystal structures ( PDB ID accession codes 3NXD and 1SVA , respectively ) . Structure inspections were performed using Swiss PDB Viewer [62] .
|
Serological studies have shown that nearly all humans are chronically infected with BK polyomavirus ( BKV ) . The infection isn't usually associated with noticeable symptoms . However , opportunistic replication of BKV in therapeutically immunosuppressed kidney transplant recipients ( KTRs ) can lead to dysfunction or loss of the engrafted kidney . BKV associated nephropathy can occur even in KTRs with high levels of anti-BKV antibodies that might be expected to neutralize the virus . In this report we provide a possible explanation: we show there are at least two BKV genotypes , which are distinct serotypes with respect to antibody-mediated neutralization . Using a novel neutralization-based approach , we found that about half of 108 KTRs did not have detectable levels of antibodies capable of neutralizing BKV genotype IV ( BKV-IV ) at the time of transplantation . Of these initially BKV-IV naïve KTRs , about half experienced acute BKV-IV specific seroconversion during the first year after transplantation . This likely reflects a de novo BKV-IV infection arising from the engrafted kidney . In a pilot study , we show that recombinant BKV-IV VLPs can induce high levels of BKV-IV-neutralizing antibodies in vaccinated animals . Our results suggest that administration of a BKV VLP-based vaccine to prospective KTRs might protect against the development of opportunistic BKV replication .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"renal",
"transplantation",
"virology",
"biology",
"microbiology",
"nephrology"
] |
2012
|
Neutralization Serotyping of BK Polyomavirus Infection in Kidney Transplant Recipients
|
The beneficial contribution of commensal bacteria to host health and homeostasis led to the concept that exogenous non-pathogenic bacteria called probiotics could be used to limit disease caused by pathogens . However , despite recent progress using gnotobiotic mammal and invertebrate models , mechanisms underlying protection afforded by commensal and probiotic bacteria against pathogens remain poorly understood . Here we developed a zebrafish model of controlled co-infection in which germ-free zebrafish raised on axenic living protozoa enabled the study of interactions between host and commensal and pathogenic bacteria . We screened enteric fish pathogens and identified Edwardsiella ictaluri as a virulent strain inducing a strong inflammatory response and rapid mortality in zebrafish larvae infected by the natural oro-intestinal route . Using mortality induced by infection as a phenotypic read-out , we pre-colonized zebrafish larvae with 37 potential probiotic bacterial strains and screened for survival upon E . ictaluri infection . We identified 3 robustly protective strains , including Vibrio parahaemolyticus and 2 Escherichia coli strains . We showed that the observed protective effect of E . coli was not correlated with a reduced host inflammatory response , nor with the release of biocidal molecules by protective bacteria , but rather with the presence of specific adhesion factors such as F pili that promote the emergence of probiotic bacteria in zebrafish larvae . Our study therefore provides new insights into the molecular events underlying the probiotic effect and constitutes a potentially high-throughput in vivo approach to the study of the molecular basis of pathogen exclusion in a relevant model of vertebrate oro-intestinal infection .
Non-pathogenic bacteria associated with animal mucosa contribute to host health and homeostasis by promoting key physiological functions and by providing protection against pathogen infections [1] , [2] , [3] , [4] . This protection , induced upon stimulation of the host immune defenses or by direct bacteria-bacteria interactions , led to the concept that carefully chosen bacteria called probiotics could be introduced in natural host microbial communities to limit infection upon colonization by pathogens [5] , [6] , [7] , [8] . Clinical evidence of probiotic efficacy in treatment of gastro-intestinal disorders and allergic symptoms triggered strong interest in the identification of biological mechanisms behind these beneficial effects [9] . Study of the protective role of probiotic bacteria during host-pathogen interactions , using comparative genomics and microbiologically controlled animal models such as gnotobiotic mice , rats , rabbits and pigs , has led to significant progress [10] , [11] , [12] . However , these mammal models are often complex and have low-throughput , while practical limits hamper identification of molecular processes behind probiotic effects , a prerequisite for prophylactic or therapeutic use of probiotics against infections [12] , [13] . As an alternative to gnotobiotic mammal models , several invertebrates , including the fruit fly Drosophila melanogaster and the nematode worm Caenorhabditis elegans , have been used to study protection provided by commensal or probiotic bacteria against pathogens [14] , [15] . New models are however needed to identify , select and evaluate factors involved in probiotic effects in a more relevant vertebrate context [4] . Recently , zebrafish ( Danio rerio ) , a tropical freshwater cyprinid and successful model in vertebrate developmental biology , proved to be convenient for studying bacterial intestinal colonization and host-pathogen interactions [16] . Zebrafish have an innate immune system and develop adaptive immunity by the age of 6 weeks , and the development and physiology of its digestive tract are very similar to those of mammals [17] , [18] . Moreover , germ-free zebrafish larvae are relatively easy to obtain and the small size and easy husbandry , combined with available genetic tools , make it particularly amenable to molecular analyses both from the host and bacterial point of view [19] , [20] , [21] . Thus far , more than twenty different bacteria have been used to infect zebrafish through various infection routes , providing valuable insight into host-pathogen interactions [16] , [22] , [23] and , more rarely , commensal/pathogen interactions within controlled intestinal microbial communities [24] , [25] , [26] , [27] . Here we developed a new experimental approach for direct analysis of host and bacterial factors involved in protection provided by exogenous probiotic bacteria against pathogens . We raised axenic zebrafish larvae on axenic live food and screened a library of intestinal fish pathogens for virulent bacteria introduced in fish water and acquired by the natural route . We found that Edwardsiella ictaluri , the causative agent of catfish enteric septicemia , is responsible for rapid lethal infection . This simple read-out of premature death enabled us to carry out a secondary screen for Gram-positive and Gram-negative non-indigenous protective bacteria . We identified 3 strains robustly protecting zebrafish larvae out of 37 potential commensal probiotic bacteria . The analysis of immunological responses in larvae , which still only exhibit innate immunity , and of the outcomes of infection in pre-colonized zebrafish , demonstrated the protective role played by probiotic adhesion factors against E . ictaluri . Our in vivo model therefore provides a relevant and potentially high-throughput approach to oro-intestinal infection so as to elucidate key events underlying pathogen exclusion by probiotic bacteria .
Studies of bacterial virulence in zebrafish have mainly been performed using conventional ( i . e . non-axenic ) larvae . To investigate the molecular bases of protection by non-indigenous probiotic bacteria against incoming pathogens in a microbiologically controlled zebrafish host , we produced germ-free zebrafish larvae by sterilizing freshly fertilized eggs using both antibiotic and chemical treatments , as previously described [20] , [28] . These germ-free larvae hatched spontaneously between 3 and 4 days post-fertilization ( dpf ) and were first tentatively fed sterile autoclaved fish food powder . However , unless a large amount of powder having deleterious effects was used , as observed in [29] , this procedure simply led to insufficient access of food particles to the mouth , likely due to poor elicitation of larval hunting behavior by non-moving food particles [30] . To circumvent this limitation and at the same time maintain adequate water quality , we fed newly hatched germ-free larvae every other day with live axenic Tetrahymena thermophila , a well studied ciliate advantageously substituting for natural zebrafish zooplankton prey [31] ( Figure S1 in text S1 ) . Standard body length [32] , and growth rate of larvae fed with T . thermophila was similar in germ-free and conventionally raised larvae ( data not shown ) . This enabled us to routinely raise germ-free larvae for up to 15 dpf at 28°C , as indicated by the absence of bacterial colony forming units ( CFU ) after plating and by negative 16S-based PCR analysis of water and homogenized larvae ( data not shown ) . To raise axenic zebrafish beyond 15 dpf , we also fed larvae axenic Artemia salina nauplii from 10 dpf onwards , therefore extending the life span of axenic zebrafish up to at least 1 month , at which point they could efficiently feed on sterile food powder . However , raising zebrafish on A . salina nauplii was labor-intensive and less experimentally amenable to multiple analyses . Therefore , we used zebrafish larvae fed only Tetrahymena for the rest of this study . We reasoned that a relevant study of protective bacteria-bacteria interactions within the intestinal tract of axenic zebrafish larvae required prior identification of virulent intestinal pathogens able to infect their host via the natural route . We screened a total of 25 potential enteric fish pathogens , including 16 different species or subspecies and several different isolates of Aeromonas , Vibrio , Edwardsiella , Listonella Photobacterium and Yersinia ( see Table S1 in text S1 ) . At 6 dpf , axenic zebrafish larvae were brought in contact with each tested pathogen by immersion for 6 h in water containing bacteria adjusted to 2 . 108 CFU/ml . After 6 h , larvae were washed and transferred to individual 24-well microtiter plate containing fresh autoclaved mineral water and incubated at 28°C under sterile conditions . Sterility of control germ-free larvae subjected to mock infection was monitored throughout the experiment by plating and 16S PCR analysis ( data not shown ) . While contact with the non-pathogenic bacterium Escherichia coli MG1655 did not affect the viability of zebrafish larvae , daily monitoring of mortality upon contact with the tested pathogens enabled us to identify the channel catfish pathogen E . ictaluri as being highly pathogenic for zebrafish larvae , leading to high and reproducible mortality within 3 days after exposure ( Figure 1A ) . We also observed that three other fish pathogens caused slightly premature mortality in zebrafish larvae: Edwardsiella tarda CIP 78 . 61 , a human and fish pathogen , and two Aeromonas strains , Aeromonas hydrophila sp . dhakensis CIP 107500 and 1 out of 6 Aeromonas hydrophyla sp . hydrophyla strains ( strain CIP 103561 ) [33] , [34] ( Figure 1A ) . Whole-mount immunohistochemistry using a polyclonal antibody recognizing various Gram-negative bacteria and CFU counts recovered from freshly euthanized homogenized infected larvae allowed us to confirm that these bacteria colonized the zebrafish gut ( Figure 1BC ) . To characterize E . ictaluri infection , we first determined whether lethality towards germ-free zebrafish larvae was dose-dependent ( Figure 2A ) , and not induced by dead heat-killed E . ictaluri ( Figure 2B ) . The number of E . ictaluri bacteria recovered from freshly euthanized homogenized infected larvae increased between 1 and 3 days post infection ( dpi ) , reaching levels of up to 4 . 8×105 CFU/larva shortly before death ( Figure 2C ) . Moreover , larvae infection with increasing dose of E . ictaluri also correlated with increased larvae colonization ( Figure 2D ) . We then tested the impact of exposure to E . ictaluri on conventional larvae and found similar sensitivity to E . ictaluri infection in axenic and conventional larvae ( mean survival reduced by 4 and 5 days , respectively; Figure 2E ) , suggesting that indigenous microbial communities developed at our facilities does not protect against E . ictaluri infection and that the absence of indigenous bacteria is not the main cause of the strong virulence of E . ictaluri in axenic zebrafish larvae . In its natural host , the primary route of entry of E . ictaluri is the intestine; however , several other potential infection routes have been reported , including olfactory sinus , gills and skin injuries [35] , [36] . To determine E . ictaluri infection sites in axenic zebrafish larvae , the E . ictaluri localization in infected larvae was monitored over time by whole-mount immunofluorescence . Consistent with CFU counts , E . ictaluri immunofluorescence signals increased from 1 to 4 dpi until larval death and were mainly detected in the gut lumen and on the head underside ( Figure 3A and data not shown ) . At 3 or 4 dpi , we sometimes observed small bacterial aggregates within the lamina propria of the distal intestine , indicating that a few bacteria had crossed the intestinal barrier ( Figure 3B ) . In the series of samples from which this image has been obtained , we observed such breaches in gut epithelium in about half of the observed individuals , generally with a single event per fish; however this was not observed in all experiments . In contrast , E . ictaluri was consistently found around the mouth and/or inside multiple abscess-like lesions ( 42±20 ( mean±SD ) bacterial clusters of 10 µm or more in diameter ) located on skin surfaces from the jaw to the gill area , or within the oral cavity ( Figure 3CD , Video S1 and Video S1 caption in text S1 ) . Altogether , these results showed that E . ictaluri entry into zebrafish larvae upon exposure by immersion led to both abundant abscesses in the peri-oral area and intestinal colonization , with occasional crossing of the intestinal barrier . To study the impact of bacterial infection on host immunological responses , we monitored mRNA levels of genes encoding inflammation markers TNFα , IL-1β , IL-22 and IL-10 , including pro- and anti-inflammatory cytokines , in axenic and infected zebrafish larvae at 1 , 2 and 3 dpi . Using quantitative RT-PCR , we observed that , while the levels of transcripts for all four cytokines remained low in axenic zebrafish larvae and in larvae exposed to the non-pathogenic bacteria E . coli MG1655 , they were higher and increased significantly over time in larvae infected by E . ictaluri ( Figure 4A–D ) . This increase required live bacteria and not only their epitopes , since incubation with heat-killed E . ictaluri did not induce inflammation ( Figure 4A–D ) nor did it reduce lifespan of larvae ( data not shown ) . A similar analysis performed with larvae infected by the 3 other milder pathogens identified in our screen ( E . tarda , A . hydrophila sp dhakensis , A . hydrophila sp hydrophila ) also revealed that they marginally induced cytokine transcripts ( Figure S2 in text S1 ) . Consistent with the localization of E . ictaluri in infected larvae , whole-mount in situ hybridization at 3 dpi revealed clusters of il1b-expressing cells in the head region , especially in the gill arches and next to localized skin lesions ( Figure 4E ) . The localization of neutrophils was also monitored over time during colonization using mpx:gfp transgenic zebrafish larvae [37] . At 3 or 4 dpi , while neutrophils were found distributed throughout the body in germ-free zebrafish and in larvae colonized by control strain E . coli MG1655 , larvae infected with E . ictaluri displayed strong neutrophil recruitment to the peri-oral region ( Figure S3 in text S1 ) . Unexpectedly , enterocytes were also seen to express GFP in E . ictaluri-infected fish , but this did not hamper identification of neutrophils . In contrast , mild pathogens did not induce significant neutrophil recruitment in the head and gut ( Figure S3 in text S1 ) . These observations were confirmed in wild-type larvae stained with Sudan black , a dye that specifically labels neutrophil granules ( data not shown ) [38] . We hypothesized that larval mortality following E . ictaluri infection could be used as a phenotypic read-out to reveal protective effects provided by known probiotic bacteria . For this , we pre-colonized freshly hatched ( 4 dpf; see Figure S1 in text S1 ) axenic zebrafish larvae with 37 commensal or probiotic Gram-positive and Gram-negative bacteria often used as probiotic strains in the food industry and/or aquaculture , including several E . coli , Lactobacillus spp . , Pediococcus spp . , Pseudomonas , Phaeobacter , Aeromonas and Vibrio strains ( see Table S2 in text S1 ) . These pre-colonized larvae were then infected at 6 dpf with E . ictaluri and we compared their mortality rate with axenic or non-infected larvae colonized only by a probiotic bacterium . This screen showed that pre-incubation with V . parahaemolyticus , E . coli ED1a-sm and E . coli MG1655 F′ , a strongly adherent and biofilm-forming commensal , provided a significant increase in survival upon E . ictaluri infection ( Figure 5A and Table 1 ) . To investigate possible direct interactions between E . ictaluri and the three identified protective strains , we first showed that in vitro exposure to E . coli MG1655 F′ or E . coli ED1a-sm bacterial-free supernatants did not impair E . ictaluri growth nor biofilm formation in microtiter plate assays ( Figure S4AB in text S1 ) . By contrast , V . parahaemolyticus supernatant slightly reduced E . ictaluri growth of ( Figure S4B in text S1 ) . Consistently , while broth co-cultures with E . coli MG1655 F′ or E . coli ED1a-sm did not reduce E . ictaluri cfu count compared E . coli MG1655 strain , co-culture with V . parahaemolyticus reduced E . ictaluri growth rate , suggesting a potentially distinct protection mechanism for V . parahaemolyticus ( Figure S4C in text S1 ) . We then compared transcription levels of il1b , tnfa and il10 in larvae pre-incubated with the most protective strain ( E . coli MG1655 F′ ) , infected or not by E . ictaluri . Whereas no inflammation could be detected in larvae colonized only by E . coli MG1655 F′ , all markers were still strongly induced upon E . ictaluri infection of pre-incubated larvae despite observed protective effects ( Figure 5B ) . However , pre-colonization with V . parahaemolyticus induced some cytokine gene expression , suggesting potential differences in the mechanisms of action of the various protective strains identified ( data not shown ) . This difference was also observed when studying the distribution of a neutrophil population of E . ictaluri-challenged larvae when pre-colonized or not by probiotic strains ( Figure 6A ) . Total counts of visible neutrophils did not significantly change under the different conditions tested , but were lower in germ-free animals . However , we found a significant redistribution of neutrophils to the head and gut at the expense of hematopoietic tissues in germ-free larvae infected by E . ictaluri ( Figure 6B and S5 in text S1 ) . Similar redistribution was also found in larvae pre-colonized by E . coli MG1655 F′ or E . coli ED1a-sm and infected by E . ictaluri ( Figure 6B ) . In contrast , larvae pretreated with V . parahaemolyticus and infected with E . ictaluri display neutrophil distribution similar to that of uninfected larvae , except for a reduction in hematopoietic tissues ( Figure 6B ) . These results further suggested that mechanisms of protection against E . ictaluri infection differ between protective strains . To specifically quantify infection with E . ictaluri , we developed a qPCR-based assay from DNA extracted from entire larvae . This assay did not show significant variation in the level of E . ictaluri colonization in germ-free vs MG1655 or MG1655 F′ precolonized larvae ( Table S4 ) . The distribution of bacteria in pre-colonized larvae challenged with E . ictaluri was assessed by whole-mount immunohistochemistry using a polyclonal antibody recognizing various Gram-negative bacteria . Although this antibody does not discriminate between protective bacteria and pathogens , abscesses were consistently observed in the peri-oral region , suggesting that protective bacteria did not impair infection of the head by E . ictaluri . By contrast , while no crossing of the gut barrier by E . ictaluri was observed when larvae were pretreated with the protective strains , we could not reach definitive conclusions regarding gut infection . Our results indicated that reduced E . ictaluri virulence by E . coli MG1655 F′ did not result from direct toxicity , nor from a change in the zebrafish inflammatory response . E . coli MG1655 F′ is a highly adherent derivative of MG1655 that carries the F conjugative plasmid and expresses F pili involved in conjugation and biofilm formation [39] . Since zebrafish larvae pre-incubated with wild type E . coli MG1655 only poorly protected against E . ictaluri infection ( Figure 7A ) , this suggested that the protective effect of by E . coli MG1655 F′ might stem from changes induced by the F plasmid . Moreover , we showed that MG1655 F′ was able to colonize zebrafish larvae better than wild-type MG1655 ( Figure 7B ) , indicating that protection of E . ictaluri infected larvae was correlated with the ability of MG1655 to colonize zebrafish , both in axenic and conventional larvae ( Figure S6 in text S1 ) To further elucidate the mechanism of this protection in MG1655 background , we used an F plasmid carrying a conjugation-deficient traD mutant that still produces the F pili adhesin ( Table 1 ) . We found that this mutant continued to increase the life expectancy of E . ictaluri-challenged zebrafish larvae , indicating that the protective function is independent of conjugation events ( Figure 7A ) . In addition , introduction of an F plasmid in the protective E . coli ED1a-sm did not significantly increase protection of pre-incubated larvae against E . ictaluri infection ( p = 0 . 07 ) , potentially due to the already strong ability of ED1a-sm to colonize zebrafish larvae compared to E . coli MG1655 ( Figure 7A and Table S3 in text S1 ) . E . coli MG1655 wild type has several adhesion factors shown to increase bacterial attachment to various surfaces , including type 1 fimbriae , curli and antigen 43 [40]–[42] . To determine whether bacterial adhesion could be a key molecular determinant in MG1655 F′ protection against E . ictaluri , we tested whether increased bacterial adhesion correlated with increased protection . For this , we pre-incubated axenic larvae with E . coli derivatives constitutively expressing different adhesins previously shown to increase bacterial adhesion to various surfaces , such as antigen 43 ( Ag43 ) , type 1 fimbriae and curli [40]–[42] . Monitoring of bacterial colonization at 9 dpf in homogenized larvae pre-incubated with these MG1655 derivatives showed that these strains did not show increased ability to colonize axenic larvae when compared to the MG1655 wild type ( Figure 7B ) . Consistently , these strains did not further delay E . ictaluri infection when compared to the MG1655 wild type ( Figure 7C ) . However , deletion of type 1 fimbriae operon genes showed that E . coli MG1655 Δfim was no longer able to protect against E . ictaluri infection ( Figure 7C ) . Type 1 fimbriae were involved in adhesion to intestinal and epithelial cells in different E . coli strains such as K1-type E . coli [43] , avian pathogenic E . coli [44] , enteroaggregative LF82 [45] and the probiotic Nissle strain [46] . Our results therefore suggested that type 1 fimbriae , and potentially other E . coli adhesins , could contribute to zebrafish tissue adhesion , reaching its maximum under wild-type conditions , since overexpression did not lead to further protection and colonization ( Figure 7BC ) . Altogether , these data indicate that E . coli MG1655 F′ adhesion capacity provided by the F-plasmid and to a lesser extent type 1 fimbriae , is involved in protection against E . ictaluri infection .
Over a century ago , Elie Metchnikoff postulated the existence of the probiotic effect [47]; however , few of its actual mechanisms were experimentally demonstrated in vivo , thus severely limiting the scope of applications of probiotics in alternative anti-infectious strategies [12] , [47] . Here we developed a controlled model of axenic vertebrate colonization to study host and bacterial aspects of probiotic-based protection against bacterial pathogens acquired by a natural route of infection . We first circumvented current limitations of existing procedures to raise axenic zebrafish and we used this new protocol to study the impact of gut microbial communities on animal health and development [20] . While use of sterile fish food powder or even the absence of feeding [28] , [48] , [49] leads to rapid epidermal degeneration followed by premature death [27] , we show here that feeding axenic live food to zebrafish larvae enabled us to routinely raise gnotobiotic larvae for over a month . We hypothesize that the natural motility of Tetrahymena cells and Artemia naupli enables young larvae to more easily feed while food continuously remains in suspension . Moreover , multiplication of Tetrahymena cells on waste reduces food-based soiling of water in microtiter plate wells , thereby diminishing the requirement for frequent food supply . Finally , our procedure mimics natural feeding behavior and reduces the impact of nutritional parameters on fish development and outcome of intestinal microbe-host interactions . Although we did not systematically raise older larvae , 1-month-old axenic zebrafish appeared morphologically healthy , suggesting that axenic zebrafish can be raised over a longer period . This new and reproducible procedure therefore opens up the prospect of studying the impact of bacteria on zebrafish gut anatomy and physiology from larval to the adult stage . Zebrafish constitute an increasingly popular model for analyzing bacteria-host interactions and bacterial pathogenicity in vivo [16] , [17] , [19] , [50] , [51] , and many studies have used inoculation infection procedures such as intramuscular or intraperitoneal injection in adults or in recently hatched larvae [16] . However , in addition to some viral pathogens [52] , [53] , to our knowledge , the only models of non-adult zebrafish infection by immersion used the pathogenic bacteria E . tarda and Flavobacterium columnare [54] , [55] , [56] . Those studies used 24 h immature embryos which did not yet have an open digestive system , and revealed modest infection efficiency and , when assessed , high variability . Hence , a critical advantage of our approach over these other models of infection by immersion is the high rate of disease incidence , allowing the use of manageable number of animals to reach statistical significance . Here , axenic zebrafish larvae were placed in contact with a large panel of pathogens and probiotics by mere immersion in bacteria-containing water . This colonization procedure led us to identify several pathogenic bacteria , including E . ictaluri , a virulent strain rapidly deadly toward axenic and conventional zebrafish larvae . E . ictaluri is an enterobacterium identified as the causative agent of enteric septicemia in channel catfish ( Ictalurus punctatus ) ; it causes substantial economic losses , affecting most fish farms and ponds in the United States [57] . While epizootic diseases associated with acute septicemia caused by E . ictaluri have been observed only in channel catfish , this bacterium was also recovered from several other fish , including the madtom Noturus gyrinus , the Vietnamese freshwater catfish Pangasius hypophtalmus , the walking catfish Clarias batrachus , the green knifefish Eigemannia virescens and the Bengal danio Danio devario , and was also shown to be highly pathogenic when injected into adult zebrafish [58] . The relevance of our zebrafish model is further underlined by the observation that the E . ictaluri 93–146 virulent catfish isolate used by Karsi et al . [59] also led to high zebrafish mortality in our model , while its non-virulent derivative 65ST turned out to be non-pathogenic ( data not shown ) [60] . Fish pathogens generally enter into their host through the gills , skin and gastrointestinal tract , and the integrity of these physical and immunological barriers determines the outcome of host-pathogen interactions [61] . Although the natural infection route of E . ictaluri in its natural hosts is poorly characterized [35] , [36] , [59] , we observed the early appearance of head cutaneous ulcers in the ventral head and lesions of intestinal tissue at later stages of infection ( 3 to 4 days post-infection ) , prior to larval death . While we can only but speculate about causes of fish mortality , use of the neutrophil reporter zebrafish line mpx::gfp suggests one possible scenario . Indeed , at an advanced stage of infection , neutrophils have relocalized from hematopoietic tissue to head and gut sites of infection . However , data obtained during early infection also suggested that neutrophils migrate first to the head only , therefore potentially leaving the gut with a transient weakening of the immune barrier against E . ictaluri intrusion ( data not shown ) . This hypothesis is consistent with predominant expression of il1b in the head , as seen by in situ hybridization . In addition to mortality induced upon E . ictaluri infection , we found that milder fish pathogens identified in our study–E . tarda , A . hydrophyla sp . hydrophyla , and A . hydrophila sp . dhakensis–also had an immunological impact upon infection of zebrafish larvae . Although this relatively small number of identified pathogens among previously described fish pathogens could be a consequence of host-specificity , we cannot exclude the possibility that some of the tested pathogens induced milder effects undetected in our stringent phenotypic screen . We used E . ictaluri lethal infection as a phenotypic read-out to investigate potential protection provided by commensal or probiotic bacteria . Several mechanisms have been proposed to explain beneficial probiotic effects , including stimulation of the host immune system , production of antimicrobial compounds or competition for the attachment site or nutrients [5] . However , few of these mechanisms were actually shown to occur in vivo [12] . Here we show that pre-colonization of axenic zebrafish by E . coli MG1655 F′ , E . coli ED1a-sm and V . parahaemolyticus protected the host against E . ictaluri . While many human probiotics , including several Lactobacilli , were tested in our study , none showed significant protective effect against E . ictaluri infection , possibly due to the aerobic nature of microbial communities hosted by zebrafish larvae . Whereas no in vitro or in vivo growth or colonization inhibition of E . ictaluri by the protective E . coli strains could be detected , we showed that V . parahaemolyticus impaired growth in broth co-cultures suggesting a distinct protection mechanism potentially involving contact-dependent toxicity or interference against E . ictaluri . Furthermore , monitoring of cytokine gene expression in infected zebrafish larvae pre-incubated or not with protective strains showed no attenuation of the zebrafish inflammatory response induced upon E . ictaluri infection . Although we may have missed modulation of other markers , the tested genes ( tnfa , il1b , il10 , il22 ) represent major actors in inflammatory responses and cover a variety of functions . In mammals , TNFα and IL1β constitute classical pro-inflammatory cytokines , known to activate leukocytes and endothelial cells among other cell types . IL-10 is also a well-known inflammation marker , but with pleiotropic anti-inflammatory functions . IL-22 is a more recently discovered cytokine that plays a protective role in bacterial infections by signaling to non-immune cells only , such as epithelial cells of the gastrointestinal tract and skin [62] . In contrast , enhanced colonization and the life expectancy of infected larvae in the presence of the biofilm-forming E . coli MG1655 F′ strain suggest that strong adhesion promoted by F pili could lead to E . ictaluri exclusion . While this exclusion could be due to direct competition upon E . coli MG1655 F′ adhesion to zebrafish intestinal tissues , other mechanisms could contribute to the observed protection effect , including alteration of tissue architecture , or modification of E . ictaluri behavior in pre-colonized larvae . Hence , this suggests that the non-pathogenic E . coli MG1655 F′ strain or engineered derivatives could be used as potential probiotic strains against E . ictaluri or its closely related human pathogen E . tarda [63] , [64] . Introduction of the F-plasmid into the already strong zebrafish colonizer ED1a-sm led to only a slight increase in protection of pre-incubated larvae , suggesting secondary mechanisms for a probiotic effect in ED1a-sm . In support of this hypothesis , we observed that the absence of type 1 fimbriae impaired the E . coli MG1655 ability to protect zebrafish larvae without affecting colonization . Furthermore , mechanisms by which E . coli MG1655 F′ protects zebrafish upon E . ictaluri infection might be different from those of V . parahaemolyticus , as evidenced by a differential inflammatory response and redistribution of neutrophils upon infection . In conclusion , we have developed a new and potentially high-throughput approach to investigating competitive exclusion and protection against pathogens in microbiologically controlled zebrafish . Direct experimental analysis of a protective effect in a genetically tractable vertebrate model organism should prove useful when studying host-pathogen interactions . It will contribute to improvising the molecular definition of probiotic effects and their use in preventive and curative treatments against pathogens .
All animal experiments described in the present study were conducted at the Institut Pasteur according to European Union guidelines for handling of laboratory animals ( http://ec . europa . eu/environment/chemicals/lab_animals/home_en . htm ) and were approved by the Institut Pasteur Animal Care and Use Committee and the Direction Sanitaire et Vétérinaire de Paris under permit #A-75-1061 . Bacterial strains and plasmids used in this study are listed in Table 1 and Tables S1 and S2 in text S1 . E . ictaluri cells were grown in brain-heart infusion medium at 30°C; E . tarda and Aeromonas strains were grown in tryptic soy broth ( TSB ) supplemented with 0 . 25% glucose at 30°C . Lactobacillus strains were grown at 30°C in MRS medium . All other strains were grown in LB ( lysogeny broth ) at 37°C unless indicated otherwise . When required , antibiotics were added to the medium at the following concentrations: ampicillin ( Amp , 100 µg/ml ) , chloramphenicol ( Cm , 25 µg/ml ) , kanamycin ( Km , 50 µg/ml ) , streptomycin ( Sm , 100 µg/ml ) , tetracycline ( Tc , 7 . 5 µg/ml ) . Proper formation of isolated E . ictaluri CIP 81 . 96 colonies on plates was achieved by supplementing BHI with 100 U/ml of bovine liver catalase ( SIGMA C1345 ) . Wild-type AB purchased from the Zebrafish International Resource Center ( Eugene , OR , USA ) , or their F1 offspring , and nacre ( melanocyte-deficient ) mutants were raised in our facility . Eggs were obtained by marble-induced spawning and bleached according to protocols described in The Zebrafish Book [65] . After spawning , all procedures were performed in a laminar microbiological cabinet and with single-use disposable plastic ware . Fish were kept in vented cap culture flasks or 24-well microtiter plates in autoclaved mineral water ( Volvic ) at 28°C . Fish were fed every two days with axenic T . thermophila . For experiments running over 18 days , larvae were fed from day 10 onwards with axenic Artemia salina . To avoid waste accumulation and oxygen limitation , we renewed at least half the volume of water every two days to keep young zebrafish healthy . Colonization of zebrafish mucosae by bacteria present on the surface of the chorion occurs rapidly after hatching [28] . To prevent it , freshly fertilized zebrafish eggs were sterilized by separating eggs into 50 ml Falcon tubes ( 100 eggs per tube ) and washed 3 times in 50 ml of water ( 3 min at room temperature under smooth agitation ) . Afterwards , eggs were treated with a mixture of antibiotics ( 500 µl of penicillin G: streptomycin ( 10 , 000 U/ml: 10 mg/ml GIBCO #P4333 ) , 200 µl of filtered kanamycin sulfate ( 25 mg/ml ) SERVA Electrophoresis #26899 ) and antifungal drug ( 50 µl of amphotericin B solution Sigma-Aldrich ( 250 µg/ml ) #A2942 ) for 4 . 5 h under agitation at room temperature . Then they were washed 3 times as described above . Next , they were bleached ( 0 . 003% ) for 15 min , resuspending the eggs every 3 min . Eggs were washed again 3 times in water and transferred to Petri dishes to be distributed into 25 cm3 culture flasks with vented caps containing 10 mL of water ( 15 eggs/flask ) . We monitored sterility at several points during the experiment by spotting 50 µL from each flask either on tryptic soy medium agar plates supplemented with glucose or on YPD plates , all incubated at 28°C under aerobic conditions . Plates were left for at least 3 days to allow slow-growing organisms to multiply . If a particular flask was contaminated , those fish were removed from the experiment . The absence of any contamination by microorganisms in the fish larvae was further confirmed by PCR using primers specific for the chromosomal 16S region . T . thermophila . ( i ) Stock . A gnotobiotic line of T . thermophila was maintained at room temperature in 20 ml of PPYE ( 0 . 25% proteose peptone BD Bact#211684 , 0 . 25% yeast extract BD Bacto# 212750 ) supplemented with 200 µl of penicillin G ( 10 unit/ml ) and streptomycin ( 10 µg/ml ) . Medium was inoculated with 100 µl of the preceding Tetrahymena stock . After one week of growth , samples were taken , tested for sterility on TSB-glucose and YPD plates and restocked again . ( ii ) Growth . T . thermophila were incubated at 30°C under agitation in MYE ( 1% milk powder , 1% yeast extract ) inoculated directly from stock at a 1∶50 ratio . After 24 h of growth , Tetrahymena were transferred to Falcon tubes and washed ( 1500 rpm , 10 min at 19°C ) 3 times in 50 ml of autoclaved Volvic water . Finally , Tetrahymena were resuspended , transferred to culture flasks with vented caps and conserved for 4–5 days . Artemia salina . ( i ) Rehydration . 5 ml of dehydrated cysts , conserved at 4° , were placed in a microfermentor ( provided with sterile air input as well as a medium entry and exit ) containing 40 ml of sterile PBS and left with ventilation for 1 h at 28°C . They were collected in a final volume of 14 ml of PBS . ( ii ) Decapsulation . 14 ml of rehydrated cysts were placed in a microfermentor and 16 ml of bleach ( 9 . 6% ) were added . With abundant ventilation , in several minutes , the cysts turned from brown to orange . At this point , cysts were collected with a 0 . 18 mm large sieve ( Hobby Aquaristik #21620 ) and rinsed with water until bleach was eliminated . ( iii ) Hatching . One ml of cysts was placed in a microfermentor containing 35 ml of filtered sea water ( 1 L of pyrolyzed water complemented with 20 g of “Instant Ocean” salts , 400 µl of sterile 1 M CaCl2 and 20 µL of sterile 1 M NaHCO3 ) and 20 µL of sterile NaHCO3 ( 1 M ) , then incubated for 24 to 48 h at 28°C with constant ventilation and continuous flow of sea water ( 125 ml per 24 h ) to replace evaporated medium . Artemia were recovered with the sieve and abundantly washed with sterile PBS . They were finally resuspended in 30 ml of PBS , diluted 1∶10 , counted and given to larvae ( 100 artemia/larvae ) . Bacteria were grown in suitable media at different temperatures until advanced stationary phase , then pelleted ( 7500 rpm for 10 min ) and washed once in sterile water . Bacteria were resuspended and transferred to culture flasks at a final 2 . 108 CFU/ml in 5 ml . Fish larvae were transferred to small Petri dishes to eliminate the chorions and dispatched to the different flasks ( 15 larvae per flask ) . After 6 h of incubation with the pathogen at 28°C , individual larvae were distributed into 24-well plates containing 2 ml of water and 50 µL of freshly prepared T . thermophila per well to properly monitor their individual fate ( 1 larva per well ) . We checked that E . ictaluri is not pathogenic to Tetrahymena , which are able to feed on E . ictaluri bacteria when co-incubated in fish water . Between 6 and 24 larvae were used per condition per experiment . Sterility of control germ-free larvae subjected to mock infection was monitored throughout the experiment by plating and 16S PCR analysis ( data not shown ) . Each experiment was repeated at least 3 times . The larva population was followed on a daily basis and mortality recorded . Dead embryos could be readily identified in microtiter wells as they become opaque . Dubious cases were systematically checked under a stereomicroscope and larvae were considered as dead when complete arrest of all movement , including any heartbeat was observed . Opacification of the larva was always found to follow shortly . Probiotic strains were grown for 24 or 48 h in suitable media and temperature . Bacteria were then pelleted and washed once in water . They were diluted at a final concentration of 2 . 107 CFU/ml . At 4 dpf , just after hatching , zebrafish larvae were put in contact with the probiotic strains by transferring them to probiotic-containing flasks ( 15 larvae per flask ) . At 6 dpf , larvae were transferred individually into wells of a 24-well plate containing sterile mineral water inoculated with pathogenic bacteria ( 2 . 107 CFU/ml , final concentration ) . Mortality was followed daily as above . Each experiment was repeated at least 2 times and between 12 and 24 larvae were used per condition per experiment . Zebrafish were euthanized with tricaine ( MS-222 ) ( Sigma-Aldrich #E10521 ) at 200 mg/ml . Then they were washed in 3 different baths of sterile PBS-0 . 1% Tween to remove bacteria loosely attached to the skin . Finally , they were transferred to tubes containing calibrated glass beads ( acid-washed , 425 µm to 600 µm , SIGMA-ALDRICH #G8772 ) and 500 µl of autoclaved PBS . They were homogenized using FastPrep Cell Disrupter ( BIO101/FP120 QBioGene ) for 45 s at maximum speed ( 6 . 5 m/s ) . Finally , serial dilutions of recovered suspension were spotted on plates . CFU were counted after incubation at the appropriate temperature . Biofilm assay: E . ictaluri was mixed in a 1∶1 ratio with filtered supernatants of probiotic strains and grown in 96-well microtiter plates at 28°C for 48 h . Microtiter plates were then washed 3 times with water and stained with crystal violet . Biofilm formation was quantified by dissolution of crystal violet and measurement at OD 595 nm . E . ictaluri growth in presence of probiotic supernants: E . ictaluri inoculum was mixed in a 1∶1 ratio with filtered supernatant from E . coli MG1655 , E . coli ED1a-sm , and V . parahaemolyticus and allowed to grow at 28° . OD 600 nm measurements were taken every 30 minutes . The assay was performed twice in microtiterplates , and 12 different wells were monitored for each condition . Broth co-cultures of E . ictaluri with the three identified protective strains . 3 ml of BHI medium was inoculated with E . ictaluri alone or with probiotic strain and co-cultures were incubated at 30°C with agitation . Serial dilutions of over-night resulting co-cultures were spotted on BHI+catalase plates in order to obtained isolated colonies ( E . ictaluri forms patches rather than individualized colonies in absence of catalase ) . Plates were incubated at 30°C overnight and E . ictaluri and E . coli MG1655 , E . coli ED1a-sm , and V . parahaemolyticus cfu were counted . E . ictaluri was distinguished from co-cultivated bacteria based on its characteristic yellowish colony morphotype . Three co-cultures were tested for each condition . Total RNAs from 3–5 pooled zebrafish larvae or a single larva were prepared using Tri-Reagent ( Sigma ) . Oligo ( dT17 ) -primed reverse transcriptions were done using M-MLV H- reverse-transcriptase ( Promega ) . Genomic DNA from 3 pooled larvae were prepared using DNeasy blood and tissue kit ( Qiagen ) . Quantitative PCRs were performed using Power SYBR Green PCR Mastermix on an ABI7300 thermocycler ( Applied Biosystems ) . For cDNAs , Ef1α was used as a housekeeping gene; for genomic DNA , zebrafish csf1r gene was quantified to normalize the amount of DNA . Data were analyzed using the ΔΔCt method as described in [66] . The primers are listed in table S5 . WISH was performed using standard zebrafish protocols [65] . To generate the IL-1β antisense probe , IL-1β was amplified by PCR with a T3-modified antisense primer ( see Table S5 in text S1 ) . PCR products were purified with Qiaquick PCR purification kit ( Qiagen ) and the probe was transcribed in vitro with T3 polymerase ( Promega ) . Unincorporated nucleotides were removed by purification on NucAway spin columns ( Ambion ) . Whole-mount immunohistochemistry was performed using standard zebrafish protocols [65] . Anesthetized animals were fixed overnight at 4°C in 4% methanol-free formaldehyde ( Polysciences , Warrington , PA ) in PBS . Permeabilization was performed by a 1 h treatment at room temperature with 1 mg/ml collagenase ( Sigma ) . As primary antibodies , a rabbit polyclonal antiserum that recognizes E . coli and a chicken anti-GFP polyclonal were used at a 1∶800 dilution . The secondary antibodies used were Cy3-labeled goat anti-rabbit IgG ( Jackson Immunoresearch ) diluted 1∶500 and Alexa 488-labeled goat anti-chicken ( Invitrogen ) diluted 1∶800 . The non-specific Cy3 signal observed in the tissues is also observed in the tissues is also observed with other rabbit polyclonal antisera . Nuclei were stained for 30 min at room temperature with Hoeschst 33342 at 2 µg/ml ( Invitrogen ) . Imaging was performed as described in detail in [67] . Briefly , video-enhanced Nomarski/DIC images of live larvae were taken using a Nikon Eclipse 90i microscope equipped with a Hitachi HV-C20 camera and video captured on miniDV tapes; single frames were later captured using BTVpro software . Images of larvae stained by WISH were taken with a Leica MZ16 stereomicroscope using illumination from above . Images of larvae stained by immunohistochemistry were taken with a fluoIII MZ-16 stereomicroscope ( Leica Microsystems , Solms , Germany ) equipped with a DS-5Mc camera ( Nikon , Tokyo , Japan ) . Confocal images of live or fixed larvae were taken with a Leica SPE inverted confocal microscope . Confocal images of live or fixed larvae were taken with a Leica SPE inverted confocal microscope equipped with 16× ( NA 0 , 5 ) and 40× ( NA 1 , 15 ) oil immersion objectives . Images were processed with the LAS-AF ( Leica ) , ImageJ and Adobe Photoshop softwares . Statistical analyses were performed using one-way analysis of variance ( ANOVA ) with Bonferroni contrasts unless indicated otherwise . Analyses were performed using Prism v5 . 0 ( GraphPad Software ) . Zebrafish transcripts measured by RT-PCR: ef1a NM_131263 il1b BC098597 tnfa NM_212859 il22 NM_001020792 il10 NM_001020785 Genes targeted for qPCR of genomic DNA: In zebrafish: csf1r gene NM_131672 In E . ictaluri: non-coding region next to the purH gene – see genomic sequence EU285521
|
The beneficial contribution of commensal bacteria to host health led to the concept that exogenous and non-pathogenic bacteria ( probiotics ) could be used to prevent infectious disease . However , the absence of relevant experimentally tractable in vivo models severely limits our understanding of the molecular processes behind probiotic effects , therefore hampering prophylactic and therapeutic use of probiotics against infections . Here we developed a protocol to raise microbe-free zebrafish larvae fed on microbe-free live food . We placed this microbiologically controlled model in contact with known pathogens and potential probiotics to investigate molecular events underlying pathogen exclusion by probiotic bacteria . We showed that Edwardsiella ictaluri , the causative agent of catfish enteric septicemia , causes rapid death of infected larvae following exposure via the natural immersion route . We used this mortality to screen potential probiotic bacteria able to extend zebrafish survival to E . ictaluri infection and thereby identified 3 protective strains . While host immune response modulation did not contribute to protection against E . ictaluri infection , comparison of protective and non-protective strains demonstrated a key role for their adhesion factors . Our in vivo approach constitutes a relevant new model of vertebrate oro-intestinal infection and provides new insight into molecular events underlying probiotic effects against incoming pathogens .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"microbiology",
"host-pathogen",
"interaction",
"escherichia",
"coli",
"animal",
"models",
"model",
"organisms",
"bacterial",
"pathogens",
"biology",
"zebrafish",
"bacterial",
"biofilms",
"gram",
"negative",
"immunity",
"innate",
"immunity",
"microbial",
"control"
] |
2012
|
A New Zebrafish Model of Oro-Intestinal Pathogen Colonization Reveals a Key Role for Adhesion in Protection by Probiotic Bacteria
|
Skeletal muscle contraction is initiated when an action potential triggers the release of Ca2+ into the sarcomere in a process referred to as excitation-contraction coupling . The speed and scale of this process makes direct observation very challenging and invasive . To determine how the concentration of Ca2+ changes within the myofibril during a single activation , several simulation models have been developed . These models follow a common pattern; divide the half sarcomere into a series of compartments , then use ordinary differential equations to solve reactions occurring within and between the compartments . To further develop this type of simulation , we have created a realistic structural model of a skeletal muscle myofibrillar half-sarcomere using MCell software that incorporates the myofilament lattice structure . Using this simulation model , we were successful in reproducing the averaged calcium transient during a single activation consistent with both the experimental and previous simulation results . In addition , our simulation demonstrated that the inclusion of the myofilament lattice within our model produced an asymmetric distribution of Ca2+ , with more Ca2+ accumulating near the Z-disk and less Ca2+ reaching the m-line . This asymmetric distribution of Ca2+ is also apparent when we examine how the Ca2+ are bound to the troponin-C proteins along the actin filaments . Our simulation model also allowed us to produce advanced visualizations of this process , including two simulation animations , allowing us to view Ca2+ release , diffusion , binding and uptake within the myofibrillar half-sarcomere .
The rise and fall of free [Ca2+] within individual muscle cells , or the calcium transient , is measured experimentally using a Ca2+ sensitive fluorescent indicator such as furaptra or mag-fluo-4 [1] . The measured fluorescence is then used to estimate changes in [Ca2+] using a spatial averaging technique [1–4] . While this has been an extraordinarily useful tool for studying Ca2+ transients in muscle fibres , this method has limited spatial and temporal resolution [1–3 , 5] . To address these limitations and to gain further understanding of how diffusion and binding of Ca2+ occur within an individual sarcomere , researchers have developed computer simulation models that attempt to overcome the spatial and temporal limitations of direct measurement . Cannell and Allen [5] developed the first 3D multi-compartment sarcomere model to examine diffusion of Ca2+ in frog muscle . The multi-compartment half sarcomere simulation has been further developed , and refined over time by Baylor and Hollingworth [2 , 3 , 6] . In general , these Ca2+ diffusion simulations , divide the sarcomere into a 3-dimensional array of smaller compartments , and use a series of ordinary differential equations ( ODE ) to represent the diffusion , binding , and pumping reactions that occur within and between the compartments . Together , these sarcomere simulations have proven to be valuable tools for the identification of the large localized concentration gradients that exist during muscle activation , the importance of Ca2+ buffers such as ATP , and for identifying experimental errors that can arise when using calcium indicators [1–6] . As with any technique or tool , there are inherent limitations with using ODEs to power a simulation , as described by Kerr [7] and Franks [8] . While ODE simulations can be computationally efficient , simulations based on ODE’s are deterministic and describe the mass action of the system . These limitations can be problematic when used in biological systems which are intrinsically stochastic , and the volumes and ionic concentrations are small [7] . An example of an ODE calculation error can be demonstrated in a half sarcomere model where the resting concentration of Ca2+ is 50 nM [1–6] . This resting [Ca2+] translates to 22 Ca2+ ions within the half sarcomere model . If that half sarcomere is then virtually divided into 4 compartments along each of the x , y , and z axes , the ODE simulation will calculate the concentration in each compartment as 50 nM which translates to about 0 . 34 ions per compartment . Whereas in reality , some compartments would contain one or possibly two Ca2+ ions , while others would be empty . More importantly though , models based on ordinary differential equations do not simulate the diffusion of ions , but simply calculate the transfer rates between compartments and reactions that occur within each compartment . Due to this limitation , it is very difficult to develop this type of simulation to be sensitive to the structural micro-architecture within the sarcomere , let alone the differences in filament distribution within that micro-architecture . As previously stated , skeletal muscle has an intricate and complex structure . The myofilament lattice is highly organized with several proteins organized into filaments which have a specific arrangement and relationship to each other within the sarcomere ( Fig 1 ) [9–11] . Additionally , the SR surrounding the myofilament is irregularly patterned [12 , 13] . While stochastic simulations of myosin and actin interaction have been developed [14] , and it has been demonstrated mathematically that the myofilament lattice produces an obstruction to the diffusion of Ca2+ ions [15] , to date a simulation of Ca2+ diffusion within the myofibril that incorporates the topological representation of the myofilament lattice or micro-architecture of the sarcomere has not been developed . It is the purpose of this paper to create a structural model of a skeletal muscle sarcomere and develop a simulation to explore the potential impact the micro-architecture of the sarcomere has on the diffusion , binding and uptake of Ca2+ ions during a single activation . This simulation will also produce visualizations of this process . Due to the topological structural complexity within the sarcomere , the sarcomere model will be limited to a 1/8th wedge of the half sarcomere . The model produced is static and does not incorporate any filament compliance [16] or cross bridge binding . The model will further be divided into 15 measurement compartments labeled MC-c0—MC-o4 ( Fig 1 ) which will allow us to count ions and reactions that occur within these compartments . We will accomplish this by first building a structural representation of the sarcomere using the software tool Cellblender , then incorporate the model into MCell software ( Monte Carlo Cell ) , as developed and described by Stiles [17] and Czech [18] . We will then run a series of stochastic simulations using MCell to simulate the release and diffusion of Ca2+ and the reactions between Ca2+ and the fixed and diffusing Ca2+ buffers ( Troponin-C and ATP ) within the sarcomere during a simulated single activation . A complete and detailed description of the model and its construction is described in the Methodology section .
We subdivided the interior of the model into 15 measurement compartments ( MCs ) . These MCs were transparent to the molecules and ions within the simulation , but enabled the measurement of molecules , ions , and reactions that occurred within their boundaries . Unlike the previous simulations of Cannell and Allen [5] and Baylor and Hollingworth [2 , 3 , 6] , in which all the compartments maintained the same volume , we divided our sarcomere model evenly so that each MC was 1/5th the length and 1/3rd the radius of the sarcomere model ( Fig 1 ) . The resulting concentric radial slices were referred to as: center ( MC-c ) , middle ( MC-m ) , and outer ( MC-o ) . The compartments were then numbered sequentially from the Z-disk , beginning with 0 and ending with 4 closest to the M-line . In contrast to previous simulations , where concentration of Ca2+ was assumed to be uniform throughout a compartment , our model allowed a continuous concentration gradient across any MC . Diffusion was simulated independent of the boundaries of the MCs . The number of ions within the measurement compartments at each time step and the fluid volume of the respective compartment were used to calculate concentration . The placement of the actin and myosin filaments within the sarcomere model displaced fluid from the measurement compartments ( MCs ) ; the corrected fluid volume of the MCs is presented in Table 1 . The completed model contained a total of 1807742 mesh faces . Simulation runs took on average 15 days to complete . Each simulation run of our model used an average of 17 . 6 E+10 random numbers , and conducted an average of 18 . 8 E+12 ray-polygon intersection tests , which result in an average of 21 . 8E+8 ray-polygon intersections . These numbers relate not only to the complexity of the simulation , but convey the number of interactions that occurred between the various elements within the simulation . d 10 = L V 2 3 × S L ( 1 ) Where: d10 = the standard measure of lattice spacing from x-ray diffraction , LV = the lattice volume , and SL = sarcomere length . The release of calcium ions through the RYRs was driven by a changing reaction rate derived from the action potential equation from Baylor and Hollingworth [6] . Peak release rate of Ca2+ into the sarcoplasm was 205 M ms-1 and the full duration at half maximum ( FDHM ) was 1 . 6 ms ( Fig 2 ) . The total number of Ca2+ ions released , represented a [Ca2+] that would equal 343 . 3 μM within the whole model if no binding or uptake occurred . In our simulations , we observed an average peak [Ca2+] of 17 . 43 ± 0 . 87 μM . The average FDHM of the change in [Ca2+] was 3 . 36 ± 0 . 16 ms . In Fig 3 , we plot the change in [Ca2+] within the whole sarcomere model over time , for seed value 10 . These values are within the ranges reported by Baylor and Hollingworth ( Δ [Ca2+] 16 . 1 μm for their model , and 17 . 5 μm for furaptra determined Δ [Ca2+] ) [6] for both experimental results , from mouse fast twitch EDL fibres at 16° Celsius , and their multi-compartment sarcomere simulation . In Table 2 . we present the average and standard deviations , for all 10 simulation runs with respect to peak [Ca2+] , and the time at which [Ca2+] peaked . It should be noted that , as our simulation was based on a structural model of the sarcomere , the fluid volume differed in the various MCs ( Table 1 ) . While we made sure that each of the MCs within longitudinal sections were of equal size , the positioning of the actin and myosin filaments within the model meant that MCs at either end had different fluid volumes . Myosin filaments originate in MC-4 , continue through MC-2 , but extend only partially into MC-1 , and do not occupy MC-0 at all ( Fig 1 ) , whereas actin filaments that originate in MC-0 , continue through to MC-3 and only extend partially into MC-4 ( Fig 1 ) . The end result is that the fluid volume is only the same in MC-2 and MC-3 , which also have the smallest fluid volume . MC-0 has the greatest fluid volume , followed by MC-1 and MC-4 , respectively ( Table 1 ) . For a single simulation run ( seed 10 ) , we plotted the [Ca2+] over time within each of the individual MCs ( Fig 4 ) . In Fig 4a , the [Ca2+] in the five outer MCs ( the MCs that are bordered by the SR and Triad ) are depicted . In Fig 4b , [Ca2+] within the middle five MCs of the sarcomere are plotted , and finally in Fig 4c , we plot [Ca2+] within the five centre MCs of the sarcomere model . When we examine the plots in Fig 4 along with Table 2 , we see that the general trend is that peak [Ca2+] within the MCs decreases and occurs later , as their distance from the release site increases . In MC-o2 , which lies under the triad , [Ca2+] peaks at 54 . 55 μM , whereas in MC-c4 which is the furthest away from the triad and positioned closest to the middle of the sarcomere , the [Ca2+] reaches only 8 . 72 μM . Further examination of Table 2 , and Fig 4 , shows a slight bias , where more Ca2+ ions have diffused towards the Z-disks and less towards the M-line . In our simulation , we saw concentration of troponin-C ( TnC ) bound with Ca2+ ( single and double bound states were grouped together ) within the whole model peak at 230 μM , with a standard deviation ( SD ) of 1 . 5 μM . These results are again very similar to the simulation results of Baylor and Hollingworth [6] . Within the individual MCs in the simulation , we are presenting our data in terms of percentage of available TnC binding sites for Ca2+ , grouping single- and double-bound states together . We have done this as the number of TnC binding sites varies between MCs , depending on where the MC intersects the actin filaments . As a result , the concentration of available TnC sites varies slightly between MCs . The MC with the lowest number of TnC binding sites is MC-4 ( directly adjacent to the M-line ) as only 2/3rd the length of the MC is occupied by actin filaments . Presenting the data in terms of percentage of available binding sites allows a more direct comparison between the MCs within the model , and is more functionally relevant . In Fig 5 , we present the percentage of TnC binding sites occupied with one or two Ca2+ ions over time for each of the MCs for seed 10 data . In MC-1 and MC-2 regardless of their radial position ( o , m , c ) ≥ 90% of the available TnC binding sites are occupied by Ca2+ ( Fig 5a ) . In MC-0 and MC-3 the percentage changes to between 80% and 90% ( Fig 5b ) , whereas in MC-4 ( closest to M-line ) the percentage of occupied TnC binding sites reaches just 60% regardless of radial position ( Fig 5a , 5b and 5c ) . A unique benefit of computer simulation is that we are able to visualize our model at any time , and examine in great detail the position and binding state of all the Ca2+ ions: free , bound to ATP , or TnC , within the model . In Figs 6 , 7 and 8 , we present still images of the simulation at peak [Ca2+] which occurs at 3 . 2 ms . We render the simulation to highlight free Ca2+ , ATP , and ATP+Ca ( Fig 6 ) , the binding state of the SERCA pumps ( Fig 7 ) , and the binding state of TnC ( Fig 8 ) . Single myosin and actin filaments are rendered within the model to assist with orientation of the sarcomere . In Fig 6 , it is easy to see how Ca2+ ions are still clustered in the MCs closest to the release site on the triad . This image also illustrates the very large amount of ATP that is present throughout the sarcomere and reveals the very large amount of Ca2+ that is buffered by ATP at the time of peak [Ca2+] . In the supplemental information we present a complete animation of the seed 9 simulation run S1 Video . In this animation , diffusion of Ca2+ , ATP , and ATP+Ca are presented with run time for the simulation . In Fig 7 , we show the SR and the SERCA pumps in the model at 3 . 2 ms , and it can be seen at this time that the pumps that are the closest to the triad release site are all completely saturated , with each SERCA pump bound with 2 Ca2+ ions . Moving towards either the Z-Disk or the M-line , we start to see SERCA pumps in these locations still in the unbound or single-bound state . In Fig 8 , we show the model highlighting TnC bound with either one or two Ca2+ ions at the time of peak [Ca2+] , 3 . 2 ms . At this time , free [Ca2+] is decreasing as the Ca2+ ions continue to be bound to buffers including available binding sites on TnC . For this reason , TnC+Ca will peak nearly 2 ms after peak free [Ca2+] at 5 . 2ms ( Fig 5 ) . It is easy to understand that at this time many of the available TnC binding locations have been bound with Ca2+ ions and that the transition between single and double bound states is very short due to the high binding constant between TnC and Ca2+ ions . In supplemental information S2 Video we present a complete animation of the simulation , highlighting Ca2+ diffusion and binding to TnC locations on the actin filaments .
When we compare the results we obtained for the continuous diffusion of Ca2+ through the sarcomere to both the spatially averaged results from experimentation and simulation results of Baylor and Hollingworth [6] , we see very good agreement between the three data sets . The results from our simulation ( Fig 3 ) closely follow both the experimental and simulation results for the change in [Ca2+] over time as obtained by Baylor and Hollingworth [6] . As our simulation model was stochastic , we ran our simulation ten times with different random number seed values , this process gave us a data set from which we were able to run some simple descriptive statistics . From the 10 simulation runs , our model returned an average peak [Ca2+] of 17 . 43 μM with a SD of 0 . 86 μM , the plot of the [Ca2+] had a FDHM of 3 . 36 ± 0 . 16 ms ( Fig 3 ) . These values are very close to the peak [Ca2+] values of 17 μM with a FDHM of 3 . 6 ms , as reported by Baylor and Hollingworth [6] for their simulation ( at length 2 . 4 μm ) , as well as their direct experimental results for change in furaptra-detected [Ca2+] of 17 . 5 μM and FDHM 4 . 5 ms ( at a length of 3 . 5 μm ) . Not only did our simulation replicate the general results , our graph of the change in average [Ca2+] over time ( Fig 3 ) matches the graphs of Baylor and Hollingworth [6] for shape and magnitude . While the results from our simulation produced a slightly larger peak [Ca2+] and slightly narrower FDHM than the simulations of Baylor and Hollingworth [6] , their results fell within the SD of our simulation results . One of the more interesting outputs from our stochastic model , was the model’s ability to generate an estimate of the SD that can be expected experimentally , this is not produced in previous ODE based models [5 , 6] , as they are deterministic . As our model is stochastic , the estimate of the SD is based on the variability of diffusion and the probability of reactions occurring within the simulation: this is a first for this type of modelling , and the variation is independent of variability in measurement . The greatest benefit of multi-compartment models is their ability to predict how Ca2+ ions will diffuse through the sarcomere and interact with various buffers and receptors at much greater resolution than is typically available in experimentation . While we were pleased that our simulation model was able to replicate the experimental and simulation results for the general [Ca2+] transients [6] , we expected the distribution of Ca2+ between the compartments of our model to differ from previous models . There are two related reasons why we expected our simulation to produce slightly different results . The first of these is that our simulation contains the myofilament structures , and we expected that the inclusion of the myofilaments would influence the diffusion pattern of the Ca2+ . As our model simulates the diffusion of Ca2+ directly , the pattern that the diffusing ions ultimately follow is influenced by collisions , interactions , and reactions within the simulation . As such , our simulation is continuous without regard for compartments , unlike previous attempts to model Ca2+ in a sarcomere . The second reason we expected our results to be different is that the compartments we used to determine how Ca2+ is diffusing and reacting within our simulation model are of equal dimensions , and not of equal fluid volume . In the mathematical models used by Baylor and Hollingworth [6] and others [2 , 3 , 5] , all compartment volumes within the simulations are equal . To create equal volumes between the concentric cylindrical compartments , the radial width of the compartments needs to decrease as you move from the central cylinder to the outer cylinder . This must occur to balance the increased circumference of the cylinder as you move outward . This results in a radially thin outer cylinder with a large circumference , and a radially thick central cylinder with a small circumference . While there are practical reasons for maintaining equal volumes between all measurement compartments in a mathematical simulation , it creates problems in a structural model , and is not necessary . In a structural model , the asymmetrical distribution of the myofilaments will naturally create asymmetrical volumes within the model . To compensate for these asymmetrical volumes the model would then require complex asymmetrical compartments to equalize the volumes . For our simulation , we chose to create our compartments using equal dimensions for radial width and length , and allowed circumference to vary by position of the cylindrical MC , knowing that the volume within each MC would change anyway depending on the unique composition of myofilament proteins within that MC ( Table 2 ) . The lack of equal volumes does not affect the diffusion or reactions of elements within our simulation though , as diffusion and reactions are simulated directly , and are independent of the MCs . It is also important to point out that within our simulation the reactions between Ca2+ and SERCA occur at the surface of the SR , where the SERCA pumps are bound , and are independent of other reactions occurring within the outer MCs . So for these reasons , we cannot directly compare the results from within the compartments of our structural model with those of previous models [3 , 5 , 6] . In our simulation model , we were particularly interested in how the inclusion of the myofilaments into our sarcomere simulation would affect the diffusion and subsequent distribution of Ca2+ following an activation . Examining how the [Ca2+] differed between the MCs in our simulation , we see support for the assertions of Shorten and Snyed [15] that the structure of the myofilament lattice impedes the diffusion of Ca2+ ions . Looking at the data in Table 2 , we see that the diffusion of Ca2+ in our simulation is biased towards the Z-disks , and is not uniform in both directions . This bias is most notable in the central and middle MCs when we compare the [Ca2+] in MC-0 and MC-3 . The triad in our model is positioned 500 nm from the Z-disk as reported by Brown [20] as well as Gomez [29] . As each MC is 230 nm in length , the triad in our model is positioned just above MC-o2 . Due to this position of the triad over MC-o2 , the changes in [Ca2+] after a simulated release are greatest in this compartment . However , when we compare the distances from the release site to the centres of the MC-0 and MC-3 compartments , we see that the centre of the MC-3 compartments are closer to the triad than the centres of MC-0 compartments . As MC-3 is closer to the release site , we would expect that the peak [Ca2+] should be higher in these compartments than in MC-0 , and that peak [Ca2+] should occur sooner . We would expect this , because diffusion should distribute Ca2+ evenly in both directions based on distance and time . We see this result only in the outer most compartments where peak [Ca2+] is higher and occurs sooner in MC-o3 than in MC-o0 . However , when we look at compartments in the middle and centre of the model , we see no difference between peak [Ca2+] , between MC-m0 and MC-m3 , or between MC-c0 and MC-c3 . This occurs even though the time to peak [Ca2+] remains as one would expect , with peak [Ca2+] occurring sooner in the compartments MC-3 ( MC-o3 , m3 , c3 ) than in MC-0 ( MC-o0 , m0 , c0 ) ( Table 2 ) . The idea that diffusion is being affected by the myofilaments , is further supported by the realization that MC-0 also has a slightly larger fluid volume than MC-3 ( Table 2 ) , and as a result more Ca2+ ions are present in MC-0 than are present in MC-3 , even though it is further from the triad ( Table 2 ) . If we consider just the number of ions present in a compartment , instead of the concentration of those ions , the effect the myofilaments have on the distribution of ions throughout the model becomes clearer ( Table 2 ) . From the result presented in Table 2 , it is obvious that the distribution of the number of Ca2+ has been shifted towards the compartments with the least number of myofilaments , and the largest fluid volume . There are 2 mechanisms at work here . i ) the myofilaments impede diffusion and ii ) due to the lower volume , the concentration rises faster with fewer ions . Both factors would slow subsequent diffusion . The concentration differences tell us that the filaments impede diffusion . This is most apparent when we compare the number of ions at peak [Ca2+] in the MCs that are closest to the Z-disk MC-o0 , MC-m0 & MC-c0 with the corresponding compartments MC-o3 , MC-m3 & MC-c3 which are situated more toward the M-line and contain more myosin filaments ( Table 2 ) . Comparing these compartments , we see in the outer compartments that MC-o0 and MC-o3 have roughly the same number of ions . However , in the middle and central sections of the model , both MC-m0 and MC-c0 have significantly more Ca2+ ions than MC-m3 and MC-c3 respectively . When we used a paired t-test to compare MC-m0 with MC-m3 , and then MC-c0 with MC-c3 , we saw that they are significantly different with a p = 1 . 27e-05 , and p = 1 . 12e-07 , respectively ( Table 2 ) . As the MCs towards the Z-disk have greater fluid volume and fewer filament structures , this seems to support the assertions of Shorten and Snyed [15] that the myofilaments can impede the diffusion of Ca2+ . The reason that this occurs in both the middle and centre MCs and not in the outer MCs would also suggest that this effect is more pronounced as distance through the MFL increases . While it could be argued that the process of diffusion will balance the [Ca2+] rather than the number of Ca2+ ions present , this balancing would occur at equilibrium which is never achieved in our simulation . In our simulation , we also measured the concentration of TnC bound with Ca2+ ( single and double bound states were grouped together ) . Within the whole model , peak TnC bound with Ca2+ peaked at 230 μM ± 1 . 5 μM at 5 . 2 ms ( 90% bound ) . These results again are very similar to the simulation results reported by Baylor and Hollingworth [6] . When we examine our simulation with respect to TnC binding Ca2+ within the MCs , we again see results that indicate that the inclusion of the myofilaments into our models affects not only the distribution of Ca2+ , but also the binding of Ca2+ to TnC . In the MC-2 compartments ( MC-o2 , MC-m2 , & MC-c2 ) which lie directly below the release site , the percentage of TnC bound with Ca2+ reaches 98% in o2 and m2 , and 90% in c2 . This peak occurs between 3—5 ms , then slowly starts to decline ( Fig 5 ) . When we then compare these results to the percentage of bound TnC in the neighbouring compartments MC-0 , MC-1 and MC-3 , we see a larger difference between these compartments than would normally be expected . Following MC-2 , the highest percentage of TnC binding occurs next in MC-1 , followed by MC-0 , which is closely followed by MC-3 ( Fig 5 ) . Peak TnC binding of Ca2+ occurs in the simulation at 5 . 2 ms . At this time , there is no difference in the percentage of bound TnC in either MC-o2 or MC-o1 , with both compartments at 98% . There is also no difference in the percentage of TnC bound with Ca2+ in either MC-o3 or MC-o0 at 80% . As in Fig 4 , we see that MC-o3 and MC-o0 have very similar percentages of bound TnC ( Fig 5 ) . This occurs in spite of the fact that MC-o0 is positioned further away from the release site than MC-o3 . The further the Ca2+ has to diffuse through the MFL , the more the diffusion rate is slowed by the MFL . This is most noticeable in the very large decrease of binding that we see in MC-4 , the series of compartments directly adjacent to the M-line in Fig 7 . In these compartments ( MC-o4 , MC-m4 , & MC-c4 ) , we see that the percentage of TnC sites bound with Ca2+ peaks at only 60% , although it appears to still be slowly increasing as our simulation ends ( Fig 5 ) . The effect that the MFL has on Ca2+ and subsequently the percentage of bound TnC , is most apparent when we consider that the triad is located at 500 nm from the Z-disk . This means that regardless of the position of the MCs in our model , the triad is positioned very close to the middle of actin filaments , which are 1055 nm long in our simulation . If diffusion of Ca2+ was not being affected by the inclusion of the MFL in our model , we would expect to see a mirror image of binding of Ca2+ to TnC , one side reflecting the other . However , the presence of the MFL in our simulation results in there being more Ca2+ bound to TnC towards the Z-disk ( MC-0 ) than towards the M-line ( MC-4 ) . This large difference in the percentage of TnC bound with Ca2+ in MC-4 that we see , is consistent in all simulations runs with different seed values , although only data from seed 10 are presented in Fig 5 . It is important to realize that binding of Ca2+ to TnC on thin filaments that do not have adjacent myosin filaments , as is the case in most of MC-0 , is inconsequential with respect to contraction . The impact of this would be expected to be greater at long SL . With our structural simulation model of a 1/8th half sarcomere , we were successful in achieving very good agreement with the experimental and simulation results of Baylor and Hollingworth [6] with respect to how average [Ca2+] changed over time after a simulated release of Ca2+ . Our simulation also reinforces that there are very large localized differences in [Ca2+] that exist within the sarcomere [2 , 5 , 6] , and that these differences exist for relatively long times , during a single activation . Also , the flexible nature of this model , and the ability to alter physical structures of the myofibril , such as the triad , SR , or MFL , will allow further exploration into the numerous disease states that alter these structures . Compared with previous simulations , our simulation introduces several novel features: the inclusion of the myofilament lattice , the actual simulation of the diffusion and reaction of Ca2+ throughout the sarcomere , and the ability to produce 3-D visualizations of the process , a potentially valuable pedagogical tool . In this study we have demonstrated that the inclusion of the MFL changes the internal volumes within the simulation , and changes how Ca2+ ions are distributed within the sarcomere during a single activation . This change in distribution of Ca2+ results in more Ca2+ being distributed in the compartments closest to the Z-disk , and a corresponding decrease in the distribution of Ca2+ towards the M-line . This change in the distribution of Ca2+ is also reflected in the way Ca2+ binds to the regulator protein , TnC , so that more TnC is bound with Ca2+ towards the Z-disk . While we were unable to include these stabilizing protein titin in the model at this time , it should be noted that titin has been demonstrated to interact with Ca2+ and is exposed in the sarcomere between the ends of the myosin filaments and the Z-disks . While we did not model exposed titin in our model , our construction of the thick filament is based on experimental measurements , as a result the volume of our modelled myosin includes the volume of titin interwoven in the thick filament . The additional benefit of this type of model , is the ability to create 3D visualizations or animations of the model at any time point of the simulation from any angle . This ability coupled with the ability to choose what processes or reactions will be visualized , gives the observer an almost unlimited number of ways to explore the data set . As a result , this type of modeling provides new insight into how Ca2+ diffuses during activation . While our simulation simplifies both the complexity and number of reactions calculated by the simulation of Baylor and Hollingworth [6] , it replicates the averaged experimental and simulation results well for a single activation . It is important to emphasize that our simulation was able to do this by modeling the diffusion of Ca2+ and reactions within the sarcomere in an entirely novel way . As computational power continues to develop , adding more complexity to this simulation model will be relatively straightforward . At this time though , our simulation model provides a good first step in this journey . We see our structural simulation model as an important step to understand how the complex micro-architecture of the sarcomere affects the diffusion , binding and uptake of Ca2+ during a single activation . In the future , we will advance this model to explore how changes in sarcomere length affect the diffusion and binding of Ca2+ as the myofilament overlap and inter-filament spacing change with sarcomere length .
Within MCell , protein structures , membranes , and organelles are described by a series of triangulated surface meshes , which represent the nano-scale environment in three dimensions . Complete and detailed descriptions of the MCell simulation algorithms are described in papers by Stiles and Kerr [7 , 30] . As in the simulations of Cannel and Allen [5] and Baylor and Hollingworth [2 , 3 , 6] , our simulation is based on a myofibril with a length of half a sarcomere . Due to the structural complexity of our model , and computational limitations , we further divided our myofibril model into a 1/8th circumference slice of the half sarcomere ( Fig 1 ) . The dimensions used in constructing the model are listed in Table 3 . The interior faces of the myofibril model are defined as reflective surfaces , with the general assumption that the number of exiting ions would balance with the number of ions that would enter . The mammalian ( represented by mouse EDL ) synthetic sarcomere was modelled to represent a whole sarcomere length of 2 . 3 μm; at this length , the actin and myosin filaments were optimally overlapped ( our actin filament was modelled at a length of 1 . 06 μm which is slightly shorter than the normal actin length of mouse EDL 1 . 16 μm [31] which has shifted our force length optimal plateau to a shorter length ) . The modelled reactions that could occur in the simulation and rates that governed them are listed in Table 4 . We subdivided the interior of the model into 15 measurement compartments ( MCs ) . These MCs were transparent to the molecules and ions within the simulation , but enabled the measurement of molecules , ions , and reactions that occurred within their boundaries . Unlike the previous simulations of Cannell and Allen [5] and Baylor and Hollingworth [2 , 3 , 6] , in which all the compartments maintained the same volume , we divided our sarcomere model evenly so that each MC was 1/5th the length and 1/3rd the radius of the sarcomere model ( Fig 1 ) . The resulting concentric radial slices were referred to as: center ( MC-c ) , middle ( MC-m ) , and outer ( MC-o ) . The compartments were then numbered sequentially from the Z-disk , beginning with 0 and ending with 4 closest to the M-line . In contrast to previous simulations , where concentration of Ca2+ is assumed to be uniform throughout a compartment , our model allowed a continuous concentration gradient across any MC . Diffusion was simulated independent of the boundaries of the MCs . The number of ions within the measurement compartments at each time step and the volume of the respective compartment were used to calculate concentration . A segment of the actin filament protein model , number 2W49 by Wu [24] , was used to develop our actin filament mesh model . Then the classic myosin S1 head protein model , 2MYS by Rayment [22] , was used as a mesh model for the myosin S1 heads on the myosin filament . There are two issues with using protein models from the RCSB-Protein Data Bank directly: the models needed to be converted into suitable mesh models for use in Blender and Mcell , and the complexity of the mesh models needed to be reduced to provide results within the maximum compute wall time of 720 hours . The final simplified protein mesh models used in our simulation maintained the volume , and shape of the original models from the RCSB-PDB , but sacrificed surface detail for compute speed . Once we had simplified , scaled , mesh models for the myosin S1 heads , we modelled the myosin filament backbone as a simple cylinder 800 nm in length ( half of a normal myosin filament ) with a radius of 9 nm . The myosin S1 heads were then paired and placed along the cylinder according to published descriptions by Zoghbi [25] and AL-Khayat [26] to create models of the myosin filament . The distal 725 nm of myosin filament ( projecting towards the Z-disk ) contained S1 heads , whereas the final length ( 75 nm ) connecting myosin to the M-line was the myosin bare zone [32] . Once this was done , the individual heads were merged with the cylinder to make one unified mesh model of the myosin filament . Myosin filaments that straddled the 1/8th dividing edge of the myofibril were sliced in half so that they would fit within the model space . Developing a model of the actin filament presented a greater challenge . The imported mesh model from the protein databank ( 2W49 ) included numerous mesh faces that were internal to the model . To eliminate this superfluous structural complexity , we created a simple cylinder of similar length and diameter ( 38 . 5 nm and 6 . 6 nm respectively ) , then deformed the mesh cylinder so that it closely matched 2W49 . The newly modelled actin segment was 99 . 8% of the volume of 2W49 by Wu [24] . Along this newly built actin segement , we selected faces on the model where the TnC were located in 2W49 , and defined these surfaces as TnC placement locations . We then defined these locations in a series of surface region files and told MCell to place individual TnC proteins at each of these sites during runtime . This was an important step , as it over-rides the default random placement algorithm in MCell . The individual 38 . 5 nm sections were then merged end to end , to develop the full actin filament length . The MFL within the myofibril is dictated by how the actin and myosin filaments are positioned relative to each other . The myofibril is isovolumetric [10] , so the spacing between the filaments increases and decreases as the sarcomere shortens and lengthens , respectively . To determine the correct filament spacing for our model at a length of 2 . 3 μm , we used the equation developed by Millman [10] that relates sarcomere length ( SL ) with lattice spacing ( LS ) ( Eq 1 ) . Using Eq 1 , and assuming a lattice volume ( LV ) for mouse muscle of 4 × 10-3 μm3 [33] , we calculated the correct myofilament lattice spacing for a SL of 2 . 3 μm and incorporated it into our model ( Eq 1 ) . The SR model is based on the descriptions and images presented by Ogata [13] . The high resolution images of the SR from Ogata [13] were wrapped around the outside surface of our myofibril model and used as a template to define the SR . This part of the model is a defined surface , populated with SERCA pumps . The triad complex was positioned 500 nm from the Z-disk , as described by Brown [20] and Gomez [29] with the terminal cisternae on either side of the T-tubule . RYR placement was confined to the surface designated as terminal cisternae ( Fig 7 ) . At runtime , MCell positioned SERCA pumps and RYR release sites randomly across their defined surfaces until the determined density was reached ( Fig 7 ) . SERCA pumps were placed at a concentration of 240 μM , as described by Fanzini-Armstrong [19] and Baylor and Hollingworth [6] , although only 60% of them were allowed to be active . Activity of RYR was assigned pseudo-randomly at runtime based on seed number . The reason for using this activity percentage was that our model did not include Mg2+ ions , and we could not use the more advanced reaction equations used by Baylor and Hollingworth [6] that simulate the competitive binding of the SERCA pump by Ca2+ and Mg2+ . For computational efficiency , we simulated Ca2+-release from the RYR by creating a reaction rate file , which dictated the rate at which Ca2+ were generated from the RYR . The values in the reaction rate file were changed every 100 microseconds to allow Ca2+ release to mimic the release equation used by Baylor and Hollingworth [6] ( Fig 2 ) . To determine if this was successful , the release rates were plotted and integrated over time . From this plot the peak release rate , the full duration at half max ( FDHM ) , and total amount of Ca2+ released was determined . These data were compared with the data of Baylor and Hollingworth [6] . The cellular micro-architecture was built using Cell-Blender tools version RC 1 . 0 , in Blender 2 . 68b . Monte Carlo simulations were performed using MCell version 3 . 2 . 1 , and simulation runs were carried out on a variety of computer hardware . Initial development was completed on a 64bit Debian GNU/linux machine , running on an Intel iCore-7-3770k processor with 24 Gb of RAM . Repeated trials with different seed values were computed in parallel , on the Compute Canada Resources computer clusters Storm and Breeze , in the WestGrid resource pool ( www . computecanada . ca ) . The resulting complexity of the model required us to select a simulation step size of 0 . 6 ns to minimize the probability of missing any reaction . Reaction data were recorded every 60 ns . Simulations were run for 11 . 66E+6 steps to provide 7 ms worth of data . As the model is stochastic , 10 simulations with seed numbers 1-10 were run . One simulation run ( seed 10 ) was continued for 23 . 32E+6 steps to simulate 14 ms , and one visualization run ( seed 9 ) was completed where the position and state of all model elements were recorded every 1 μs . For data safety , a data checkpoint file was created every 6 hours so that in the event of a computer failure or shutdown , the simulation could be restarted beginning at the last 6-hour interval . The results from seed 9 were used to create the animations and 3D visualizations . The visualizations record 1 frame of the model every 5000 iterations , so the final animation takes a reasonable time to view . This does create the illusion in the animation that the ions take much larger steps in their random walk than they actually do . Raw data files from the results were processed and analyzed with Matlab-14a using a series of custom developed functions to consolidate the large data files . Matlab was used to plot the results , and compile general statistics such as standard deviations and means . Means and standard deviations were calculated for the whole model as well as the MCs within ( S1 Video ) . Paired T-tests were used to compare pairs of MCs within the model to determine how [Ca2+] , peak number of Ca2+ ions , and the time at which peak [Ca2+] occurred . Alpha was set at 0 . 01 . The Paired T-test was used as one compartment was directly compared with another compartment; only two comparisons were made .
|
In this study we develop a structural stochastic diffusion model , to study how calcium ions diffuse and interact within a skeletal muscle sarcomere following a simulated muscle activation . This type of model allows us to explore how structural elements , namely the actin and myosin filaments , within the sarcomere affect calcium diffusion , and the model provides a level of resolution for ligand-protein reactions that is currently unattainable with other types of diffusion models . This type of model is novel in that we topologically represent common protein structures within the sarcomere model . This topological representation of the myofibril includes the actin and myosin proteins which compose the filament lattice , the shape of the sarcoplasmic reticulum ( SR ) , the placement of the SR bound calcium pumps ( SERCA ) , as well as the position of the calcium release sites ( RYR receptors ) within the triad ( SR and T-tubule connection ) , which together describe the micro-architecture of the sarcomere . This type of modelling is also unique as it allows the results to emerge from the interaction of the defined parameters powered by the random nature of diffusion . Using this new model we are able to show that calcium ions do not diffuse in a uniform fashion from the calcium release site , instead the diffusion is biased towards z-disks where there are fewer myosin filaments . We also see evidence that previous modelling techniques may be underestimating both the release rate from troponin-c , and the transfer rate of SERCA .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methodology"
] |
[
"cell",
"motility",
"medicine",
"and",
"health",
"sciences",
"actin",
"filaments",
"myofibrils",
"muscle",
"tissue",
"simulation",
"and",
"modeling",
"molecular",
"motors",
"actin",
"motors",
"myofilaments",
"motor",
"proteins",
"research",
"and",
"analysis",
"methods",
"contractile",
"proteins",
"animal",
"cells",
"proteins",
"biological",
"tissue",
"muscle",
"cells",
"biochemistry",
"cytoskeletal",
"proteins",
"biochemical",
"simulations",
"sarcomeres",
"cell",
"biology",
"anatomy",
"myosins",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"computational",
"biology"
] |
2019
|
A stochastic simulation of skeletal muscle calcium transients in a structurally realistic sarcomere model using MCell
|
Laboratory confirmation of Zika virus ( ZIKV ) infection during pregnancy is challenging due to cross-reactivity with dengue virus ( DENV ) and limited knowledge about the kinetics of anti-Zika antibody responses during pregnancy . We described ZIKV and DENV serological markers and the maternal-fetal transfer of antibodies among mothers and neonates after the ZIKV microcephaly outbreak in Northeast Brazil ( 2016 ) . We included 89 microcephaly cases and 173 neonate controls at time of birth and their mothers . Microcephaly cases were defined as newborns with a particular head circumference ( 2 SD below the mean ) . Two controls without microcephaly were matched by the expected date of delivery and area of residence . We tested maternal serum for recent ( ZIKV genome , IgM and IgG3 anti-NS1 ) and previous ( ZIKV and DENV neutralizing antibodies [NAbs] ) markers of infection . Multiple markers of recent or previous ZIKV and DENV infection in mothers were analyzed using principal component analysis ( PCA ) . At delivery , 5 . 6% of microcephaly case mothers and 1 . 7% of control mothers were positive for ZIKV IgM . Positivity for ZIKV IgG3 anti-NS1 was 8 . 0% for case mothers and 3 . 5% for control mothers . ZIKV NAbs was slightly higher among mothers of cases ( 69 . 6% ) than that of mothers of controls ( 57 . 2%; p = 0 . 054 ) . DENV exposure was detected in 85 . 8% of all mothers . PCA discriminated two distinct components related to recent or previous ZIKV infection and DENV exposure . ZIKV NAbs were higher in newborns than in their corresponding mothers ( p<0 . 001 ) . We detected a high frequency of ZIKV exposure among mothers after the first wave of the ZIKV outbreak in Northeast Brazil . However , we found low sensitivity of the serological markers to recent infection ( IgM and IgG3 anti-NS1 ) in perinatal samples of mothers of microcephaly cases . Since the neutralization test cannot precisely determine the time of infection , testing for ZIKV immune status should be performed as early as possible and throughout pregnancy to monitor acute Zika infection in endemic areas .
Zika virus ( ZIKV ) is an arthropod-borne flavivirus closely related to several human pathogens of public health significance , including dengue viruses ( DENV1-4 ) , yellow fever virus ( YFV ) , Japanese encephalitis virus ( JEV ) and West Nile virus ( WNV ) [1 , 2] . The emergence of ZIKV in the Pacific Islands and more recently in the Americas has been linked to an unprecedented range of neurological disorders and congenital syndrome , notably Guillain-Barré syndrome and fetal microcephaly [3–6] . Currently , autochthonous transmission of ZIKV has been reported in over 50 countries [7] . The countries most affected by extensive ZIKV outbreaks overlap with geographic areas where dengue has been endemic or hyperendemic [8 , 9] . Thus , a significant proportion of the population living in these regions has already been exposed to at least one DENV serotype [9] . ZIKV and DENV share a high degree of sequence identity and structural similarity [10] , leading to potential immunological cross-reactivity [11 , 12] . To address the challenges in serology-based testing of flavivirus-immune patients , a considerable amount of effort has been made to establish reliable serological assays with the ability to discriminate ZIKV and DENV infections [13 , 14] . However , serological assays based on the detection of binding antibodies are difficult to interpret due to cross-reactivity , and virus neutralization tests need to be performed to rule out misleading positive serological results , especially in areas where past/current DENV circulation has been recorded [14–19] . The plaque reduction neutralization test ( PRNT ) measures a functional subset of virus-specific neutralizing antibodies ( NAbs ) and , consequently , is more specific than antibody binding assays [20 , 21] . PRNT remains the gold standard test to distinguish between previous flavivirus infection and to discriminate different DENV serotypes . However , virus neutralization assays are time-consuming , and appropriate laboratorial infrastructure and skilled technical staff are required to perform the tests [14 , 20 , 21] . Moreover , transient cross-neutralization might also contribute to difficult result interpretation when testing samples collected during or soon after recovery from individuals experiencing a secondary flavivirus infection [15 , 16 , 18 , 19] . Brazil experienced the largest ZIKV epidemic in the Americas and was the first country to report an increase in the prevalence of microcephaly associated with ZIKV infection [22] . The northeast region , the second most populous area of Brazil , accounted for approximately one third of the notified ZIKV cases [23] and for the vast majority of the microcephaly cases reported in the country [24] . The reasons for the higher incidence of ZIKV infections and microcephaly in this region remain unclear . Additionally , the population exposure to ZIKV after the first outbreak in the northeast has not yet been fully explored . We have previously reported a case-control study showing a causal link between microcephaly at birth and congenital ZIKV infection [25] . Moreover , we demonstrated a high frequency of ZIKV-specific NAbs on the mothers of cases and controls [25] . Here , well-characterized samples of this case-control study were used to explore in-depth ZIKV and DENV NAbs profile among the mothers of cases and controls according to their previous dengue exposure . In addition , we described the dynamics of the transplacental transfer of ZIKV and DENV antibodies .
This study was part of a case-control study carried out in the city of Recife ( Pernambuco state , Northeast Brazil ) , between January and December 2016 [25] , after the peak of the microcephaly epidemic ( November 2015 ) in this region [24] . DENV was first detected in Pernambuco State in the early 1980s [26 , 27] , while autochthonous transmission of Zika and chikungunya viruses was reported in early 2015 [28] . Yellow fever and dengue vaccines are not recommended or publicly available in this setting . Since 2017 , transmission risk of YFV has been a concern in some Brazilian states due to clustered outbreaks , mainly in the South and Southeast regions , where YFV vaccination is currently recommended for residents and travelers . Pernambuco state has not been included in the current YFV risk areas by the Pan American Health Organization ( PAHO ) [29] . Details of the study design , inclusion/exclusion criteria and data collection of the case-control study have been previously described elsewhere [25] . Cases were defined as newborns ( alive or deceased ) with head circumference of 2 standard deviations ( SD ) below the mean for sex and gestational age in the appropriate chart . Controls were live newborns without microcephaly and with normal brain imaging by transfontanellar ultrasonography ( USG ) . For each microcephaly case , two controls were selected and matched by area of residence and expected date of delivery . Mothers and their newborns were consecutively recruited at birth in eight public maternity units in the metropolitan region of Recife . Mothers were interviewed by trained nurses using a structured standardized questionnaire . We included 89 out of 91 mothers of cases ( two mothers without biological specimens ) and 173 controls . Maternal blood samples were drawn solely at the time of admission for delivery , and no mothers reported febrile illness . For all neonates , umbilical cord blood was collected , and cerebrospinal fluid ( CSF ) was collected only for cases . These samples were collected immediately after birth . Serum samples were separated and stored at -70°C until tested . For the mothers , we considered recent ZIKV infection if the subject tested positive with a ZIKV RNA test using real-time quantitative reverse transcription polymerase chain reaction ( qRT-PCR ) and/or for ZIKV-specific immunoglobulin M ( IgM ) and/or immunoglobulin G subclass 3 ( IgG3 ) anti-NS1 ( nonstructural protein 1 ) antibodies by enzyme-linked immunosorbent assay ( ELISA ) . Positivity for ZIKV-specific IgG3 anti-NS1 indicates extended recent infection herein interpreted as recency of infection ( IgG3 anti-NS1 antibodies are detectable for 4 to 6 months after infection [30] ) . Previous exposure to ZIKV and/or DENV exposure was confirmed if mothers tested positive for ZIKV and/or DENV1-4 NAbs by PRNT . In addition , the following classifications were used: ( i ) naïve population , NAbs for ZIKV or DENV were undetectable ( antibody titers <1:20 ) ; ( ii ) single flavivirus infection , mothers tested positive for only one flavivirus infection ( antibody titers >1:20 for only one virus ) ; ( iii ) multiple flavivirus infections , characterized by the detection of NAbs ( titers >1:20 ) for more than one virus . We defined laboratory-confirmed microcephaly neonates as those who tested positive for the detection of the ZIKV genome by qRT-PCR and/or ZIKV-specific IgM by ELISA in any biological specimen ( umbilical cord blood and/or CSF ) at birth . Written informed consent was obtained from a parent or guardian of each parturient or healthy infant enrolled in the study . The protocol for the study was approved by the Research Ethics Committee of the Pan American Health Organization ( PAHO-2015-12-0075 ) and Aggeu Magalhães Institute , Oswaldo Cruz Foundation ( IAM/ FIOCRUZ-PE ) ( CAAE: 51849215 . 9 . 0000 . 5190 ) . Serum samples of mothers and neonates ( cases and controls ) and CSF of neonates ( cases ) were tested by qRT-PCR for the detection of the ZIKV genome and IgM capture ELISA for the detection of ZIKV-specific IgM antibodies . For qRT-PCR , virus RNA was extracted using the QIAamp Viral RNA Mini Kit ( Qiagen , Valencia , CA ) following the manufacturer’s instructions . One-step qRT-PCR was performed using primers and probes described by Lanciotti and colleagues [15] . IgM-capture ELISA was conducted using a protocol previously described [31] and reagents were provided by the US Centers for Disease Control and Prevention ( CDC; Fort Collins , CO , USA ) . To account for potential flavivirus cross-reactivity , samples were tested in parallel with ZIKV and DENV antigens . ZIKV antigen ( CDC Vero E6–derived , inactivated ZIKV antigen ( whole virus ) ; virus strain H/PF/2013 ) and normal antigen ( CDC Vero E6–derived , mock-infected normal antigen ) were kindly provided by CDC . DENV antigen , a mixture of the 4 DENV serotypes , was prepared in a similar fashion using virus strains isolated in the study setting ( Recife , Northeast Brazil ) : DENV-1 ( BR-PE/97-42735 ) , DENV-2 ( BR-PE/95-3808 ) , DENV-3 ( BR-PE/02-95016 ) , and DENV-4 ( BR-PE/12-008 ) . Positive ( CDC humanized 6B6C-1 pan-flavivirus ) and negative ( pooled flavivirus-negative serum ) controls were included in each plate . All sera were tested in duplicate , and the results were calculated as a ratio of the average optical density ( OD ) value of the test sample ( P ) divided by the average OD value of the negative control ( N ) . P/N values of <2 . 0 were considered negative; >3 . 0 , positive; and 2 . 0–3 . 0 , equivocal . Samples showing positive results for both Zika and dengue antigens were considered positive for ZIKV IgM only if the ZIKV P/N ratio was at least twice the DENV P/N ratio [31] . Maternal sera were additionally tested by a novel in-house ELISA for the detection of ZIKV-specific IgG3 anti-NS1 antibodies , following a protocol described in details elsewhere [30] . Serum samples , in duplicate , were tested in parallel using purified NS1 proteins from Zika and DENV1-4 as antigens . Recombinant NS1 proteins expressed in the mammalian cell line 293 ( Native Antigen Company , Oxfordshire , UK ) included ZIKV NS1 ( strain Uganda/MR/766 ) and DENV-1 ( strain Nauru/Western Pacific/1974 ) , DENV-2 ( strain Thailand/16681/84 ) , DENV-3 ( strain Sri Lanka D3/H/IMTSSASRI/2000/1266 ) and DENV-4 ( strain Sri Lanka D3/H/IMTSSA-SRI/2000/1266 ) NS1 . Assay controls included: Zika positive ( sera from convalescent patients collected 60 days post onset of symptoms ) ; dengue positive recent infection ( pooled sera from early convalescent virologically and/or serologically confirmed dengue patients , collected 20–30 days post onset of symptoms ) ; and flavivirus-naïve sera ( human type AB serum from healthy individuals from USA [MP Biomedicals , Solon , USA] diluted in IgG depleted human serum [Molecular Innovation , Novi , USA] ) . For ZIKV IgG3 analysis , the results were calculated as a ratio by dividing the average OD value of the test sample by the average OD value of the dengue recent infection control . The cut-off value for ZIKV IgG3 anti-NS1 antibodies positivity was based on a ratio>1 . 2 [30] . Positive ZIKV IgG3 anti-NS1 results were interpreted as recent ZIKV infections . ZIKV and DENV1-4 NAbs were assessed by PRNT in maternal and neonate sera following a standardized protocol [30–32] . Briefly , PRNT was conducted in Vero cells seeded at a density of 300 , 000 cells/mL using 24-well plates . Serum samples were heat-inactivated ( 30 minutes at 56°C ) , serially diluted ( 4-fold dilution , starting at 1:20 ) and mixed with ≈30–100 plaque-forming units ( PFU ) of each challenge virus . Virus strains used in the assay were isolated in the study setting ( Recife , Northeast Brazil ) : ZIKV ( BR-PE243/2015 ) , DENV-1 ( BR-PE/97-42735 ) , DENV-2 ( BR-PE/95-3808 ) , DENV-3 ( BR-PE/02-95016 ) , and DENV-4 ( BR-PE/12-008 ) [31 , 32] . The virus-serum mixtures were incubated for 1 h at 37°C and then transferred to a monolayer of Vero cells to allow virus adsorption . After incubation , cells were covered with semisolid medium and incubated for 6–7 days at 37°C . Next , the cell monolayer was fixed with formalin solution and stained with crystal violet . The plates were washed and allowed to dry before plaque counts . The cutoff value for PRNT positivity was defined based on a 50% reduction in plaque counts ( PRNT50 ) in the lowest dilution tested ( 1:20 ) . Final titer calculation does not take into account the dilution made by mixing virus and antibody dilution ( 1:1 volume ratio ) . To estimate virus-specific NAb titers ( log 10 transformed ) , we calculated IC50 values by nonlinear regression using the sigmoidal dose response ( variable slope ) equation on GraphPad Prism 7 . 0a . Serum samples were considered positive for each virus tested when antibody titers were >1:20 dilution . The PRNT assay for ZIKV has been validated using a well-characterized panel of longitudinal serum samples from primary and secondary flavivirus-infected individuals collected before and after ZIKV introduction in Brazil [30] . All laboratory procedures were conducted at the Virology Department ( LaViTe ) of the Aggeu Magalhães Institute , Oswaldo Cruz Foundation ( IAM/ FIOCRUZ-PE ) . Statistical analysis was performed using SPSS software ( version 12 ) and Graph Pad Prism ( version 7 . 0a ) . We described the frequency of recent ZIKV infection and ZIKV and dengue serotype-specific ( DENV1-4 ) NAbs among mothers of cases and controls . Serological profiles among mothers of cases and controls were analyzed using a conditional logistic regression . The stratified analysis of the association between maternal ZIKV positivity ( NAbs ) and previous dengue exposure was performed using the Mantel-Haenszel test . We applied a dot plot to show the NAbs titer distribution of mothers of cases and controls according to their previous dengue immune status . Antibody titers were compared between groups using a nonparametric Mann-Whitney test . We further explored the distribution of ZIKV NAbs titers among mothers according to two categories of microcephaly cases ( ZIKV laboratory-confirmed and laboratory-negative cases ) and the control group . For all comparisons of maternal antibody titers , we included 89 cases and 89 controls . The frequency of ZIKV exposure by these categories was analyzed by a chi-square test for linear trend . We performed principal components analysis ( PCA ) to identify patterns and simplify structures underlying the multiple markers of recent or previous ZIKV and DENV infection among mothers , to identify groups of variables that were mainly correlated with each component , and to calculate individual scores related to each one of these components . The explanatory variables used were ZIKV-specific IgM and IgG3 anti-NS1 antibodies and ZIKV , DENV-3 and DENV-4 NAbs . The Kaiser-Meyer-Olkin ( KMO ) test was used to assess the measure of sampling adequacy . Bartlett’s test of sphericity was also applied to verify the sufficiency of the correlation between the variables for the PCA analysis , where a nonsignificant result ( p>0 . 05 ) would indicate a lack of suitability of the variables for identifying underlying components . In the first step of PCA , we retained factors in the model with an eigenvalue ≥ 1 . 0 . The second step was to identify variables strongly correlated with each component . These coefficients were the factor loadings generated in the component matrix . The individual factor scores for each component were transformed to a 0 to 1 scale and then added to the dataset . The dependent variable was the neonates’ status ( microcephaly case or control ) , and the difference between the factor scores within these groups was assessed by the Wilcoxon nonparametric test for two independent samples . For the principal component analysis , we did not take into account correlation . For the analysis of placental transfer , maternal and newborn antibodies mean titers were compared using the nonparametric Wilcoxon test for paired samples . The placental transfer ratio ( TR ) was calculated as follows: TR = [newborn antibody titer/maternal antibody titer] × 100 . The median TRs of ZIKV antibodies between cases and controls were compared using the Mann-Whitney test . For the comparisons of TRs between ZIKV , DENV-3 and DENV-4 groups , we used a nonparametric Kruskal-Wallis test . Pearson correlation was used to measure the association between maternal levels and transfer ratio ( TR ) of ZIKV antibodies to the newborn . The level of significance was set at 0 . 05 .
A total of 262 mothers of 89 neonates born with microcephaly ( cases ) and 173 controls without microcephaly were analyzed . The mean maternal age was similar between the cases and control groups . At the time of delivery , no mothers had positive ZIKV RNA tests ( qRT-PCR ) . Overall , 7 . 25% of the mothers were positive for recent ZIKV infection by IgM and/or IgG3 anti-NS1 antibodies . Only 5 . 6% of the mothers of cases and 1 . 7% of the control group were positive for ZIKV-specific IgM . Considering recency of infection , 8 . 0% and 3 . 5% were positive for IgG3 anti-NS1 for mothers of cases and controls , respectively , but these differences were not statically significant . Overall , 61 . 4% of the participants had ZIKV infection ( based on NAbs ) independent of their dengue immune status . ZIKV infection detection by PRNT was slightly higher among mothers of cases ( 69 . 6%; 95%CI 59 . 4–78 . 1 ) than that of mothers of controls ( 57 . 2%; 95%CI 49 . 7–64 . 3 ) ( p = 0 . 054 ) ( Table 1 ) . Table 1 also shows the profile of flavivirus-specific NAbs for the mothers . The majority of the mothers of cases ( 80 . 9% ) and controls ( 82 . 7% ) had NAbs to multiple flaviviruses , predominately for the combination of ZIKV , DENV-3 and DENV-4 . The overall frequency of mothers with NAbs to DENV-1 and DENV-2 was lower than 3% . ZIKV was the single flavivirus infection in 7 . 8% ( 7/89 ) of the mothers of cases and 4 . 6% ( 8/173 ) of controls . Eight mothers of cases ( 8/89 ) and 14 ( 14/173 ) mothers of controls did not have detectable antibodies for ZIKV or DENV exposure . Overall , mothers with serological markers of exposure to DENV infection were more likely to have detectable levels of NAbs to ZIKV than dengue-naïve mothers ( 64 . 9% vs . 40 . 6% , respectively; RR: 1 . 60; 95%CI 1 . 07–2 . 39; p = 0 . 006 ) ( Table 2 ) . This higher frequency of ZIKV positivity among mothers exposed to dengue was also observed when analyzing by mothers of cases ( 74 . 3% vs . 46 . 7%; p = 0 . 033 ) and controls ( 60 . 2% vs . 36 . 3%; p = 0 . 034 ) . Fig 1 shows the ZIKV-specific NAbs titers among mothers of cases and controls according to their DENV exposure status . The mean titers of ZIKV NAbs antibodies between DENV-naïve and DENV-infected mothers were similar for mothers of cases ( p = 0 . 150 ) or controls ( p = 0 . 414 ) . Higher levels of ZIKV NAbs titers were detected among mothers of microcephaly neonates with laboratory confirmation of recent ZIKV infection ( qRT-PCR and/or ZIKV-specific IgM ) ( n = 31 ) when compared to that of mothers of controls without microcephaly ( p = 0 . 010; Fig 2 ) . Additionally , there was a statistically significant trend of positivity for ZIKV NAbs when considering mothers of laboratory-confirmed microcephaly cases ( OR = 2 . 84 ) , mothers of microcephaly cases not laboratory-confirmed ( OR = 1 . 22 ) and controls ( chi-square for trend: 4 . 90; p = 0 . 026 ) . Considering the PCA , the KMO test result was 0 . 530 , and Bartlett’s test was significant ( p< 0 . 0001 ) , which indicated a sufficient correlation between the variables to perform the analysis . A two-factor model emerged from the data reduction when the criteria of eigenvalues >1 . 0 were applied . This two-factor model accounted for 59% of the total variation . After varimax rotation , component 1 presented the variables strongly correlated with previous DENV infection , and component 2 presented the variables strongly correlated with recent and previous ZIKV infection ( Fig 3 ) . For each individual , we calculated the scores related to each component . For component 1 , there was no significant difference between means of mothers of cases ( mean rank: 127 . 1 ) compared to those of mothers of controls ( mean rank: 129 . 9; p = 0 . 768 ) . For component 2 , the scores of recent and past ZIKV infection were higher among mothers of cases ( mean rank: 142 . 7 ) than those of mothers of the controls ( mean rank: 121 . 9; p = 0 . 033 ) . Table 3 and Fig 4 show the placental transfer of ZIKV- , DENV-3- and DENV-4-specific antibodies in the mother-neonate pairs . ZIKV-specific NAbs levels were significantly higher in the neonates than in the corresponding mothers ( TR: 104 . 7%; p<0 . 001 ) . A similar pattern was observed for the levels of DENV-3- ( TR: 109 . 1%; p<0 . 001 ) and DENV-4- ( TR: 107 . 9%; p<0 . 001 ) specific antibodies . Dengue serotype-specific NAbs were more efficiently transferred through the placenta than ZIKV antibodies ( DENV-3>DENV-4>ZIKV; p<0 . 001 ) . The placental transfer ratio of ZIKV antibodies to the neonates did not differ between cases and controls ( p = 0 . 915; Fig 4A ) . We found a positive correlation between ZIKV antibody titers in the maternal and newborn samples ( r = 0 . 866; p<0 . 001 ) . This serological pattern was similar between cases ( r = 0 . 8994; p<0 . 001 ) and controls ( r = 0 . 8432; p<0 . 001; Fig 4B ) .
The emergence of ZIKV in the Americas and its cocirculation in dengue endemic areas has hampered the diagnosis of flavivirus infections [14] . The potential for cross-reactivity of antibodies induced by these viruses has complicated serological confirmation of Zika and dengue infections , requiring cumbersome confirmatory assays such as neutralization tests [11 , 12 , 15–18 , 20] . In this study , we used a large panel of serum samples from a case-control study to unveil the ZIKV and DENV immune profiles of mothers after the first wave of ZIKV transmission in a previously unexposed population . The findings reported inherently reflect the antibody profile at the time of delivery of mothers of microcephaly cases and controls living in the epicenter of the microcephaly epidemic , which is also a high DENV transmission area [26 , 27] . At the time of delivery , mothers had a low frequency ( ~8% ) of positivity for ZIKV-specific IgM and a novel IgG3 anti-NS1 assay , the latter measuring recency of Zika exposure in the past six months [30] . In a previous publication , our group reported 31% recent ZIKV infection ( IgM ) among mothers of microcephalic infants in the same setting [31 , 33] . One of the possible explanations for the variations in the incidence of ZIKV infection among mothers is the different microcephaly inclusion criteria between studies . In the initial publication , when the etiological cause of congenital microcephaly was under investigation , the case series included only severely microcephalic infants recruited at the early stage of the outbreak [31 , 33] . Nevertheless , both studies showed that at delivery markers of recent ZIKV infection among mothers might not be detectable . These findings support the waning of IgM antibodies over time [34] , suggesting a lag time between virus exposure and time of delivery . In fact , higher levels of ZIKV NAbs titers among mothers of microcephalic neonates with laboratory confirmation of recent ZIKV infection are consistent with a more recent infection . However , neither the case-series nor the case-control study allows for the assessment of the timing of ZIKV infection among mothers of microcephaly cases . These issues related to the timing of ZIKV infection and development of adverse outcomes during pregnancy can be further investigated in ongoing cohort studies . Notably , our findings demonstrated a strikingly high prevalence of ZIKV exposure , as defined by PRNT , in the study population close to the peak of the microcephaly outbreak in Northeast Brazil . We found that approximately 60% of the mothers had already been infected by ZIKV , suggesting a high transmission rate in a short period of time ( 2015–2016 ) after the introduction of ZIKV into this naïve population . This finding further supports the information of high exposure rates from previous Zika outbreaks in Micronesia [35] , French Polynesia [36] , Brazil [30 , 37] and other Latin America countries [38] . Notably , the low proportion of ZIKV-naïve women might explain the reduction in the number of cases of congenital Zika syndrome following the first large Zika outbreak in this setting [24] . Our results also demonstrated a high prevalence of multiple flavivirus exposures among mothers of microcephaly cases and controls at birth . Our data confirmed that DENV-3 and DENV-4 serotypes were the predominant pre-exposure dengue NAbs , as reported by other studies conducted before [32] and after [25 , 33 , 39] the ZIKV outbreak in the same setting . The predominance of DENV-3 and DENV-4 serotypes in our study population reflects the epidemiological scenario of DENV circulation in the last decade in the study setting [26 , 27] . DENV-3 was the sole serotype circulating in Recife between 2002 and 2006 . DENV-4 predominantly circulated in the setting after its introduction in 2010 . DENV-1 and DENV-2 circulated in the city in the 1990s before DENV-3 introduction . Data from different epidemiological studies conducted by our group and from the surveillance system of the city registered lower rates ( nonepidemic ) of DENV-1 and DENV-2 circulation between 2012–2014; however , these viruses were quickly displaced by the introduction and circulation of ZIKV and CHIKV in the setting [40] . Although immunity to DENV-1 and DENV-2 was present among the DENV-immune mothers , it represented only ~3% of the participants , and we concentrated our analysis on DENV-3 and DENV-4 serotypes . We acknowledge that this lower frequency of DENV-1 and DENV-2 serotypes might be related to the sensitivity of the assay to detect lower levels of NAbs ( <1:20 ) , considering that these viruses circulated predominantly in the 1990s in the study setting . However , our study population was comprised of mostly young mothers ( ~25 years old ) who had not been exposed to DENV-1 and DENV-2 outbreaks in the setting . We also acknowledge that cross-reactions might complicate PRNT results interpretation and that multiple viruses positivity might , at some extent , represent some levels of cross-reactivity . However , the pattern of DENV serotype-specific profile found in our study matches the epidemiological scenario of DENV circulation in the setting . Our finding of high frequency of previous dengue exposure among mothers is in consonance with a recently published retrospective case-control study in the neighboring state of Paraiba [41] . Notably , we found that ~10% of the mothers of microcephaly cases did not show any serological markers of exposure to DENV and ZIKV by PRNT , which may be related to the sensitivity of the assay at the serum dilution used ( 1:20 ) . We acknowledge that this frequency of ZIKV negatives might also represent mothers of microcephaly cases not associated with ZIKV infection during pregnancy . Prior to the Zika epidemic ( 2000–2014 ) , the annual prevalence of microcephaly cases was 5 . 0 ( CI: 3 . 6–6 . 6 ) per 100 , 000 live births , which corresponds to an annual average number of 44 cases in the Northeast region [42] . Considering this previous prevalence and the population sampled ( 13 , 531 live births ) , the expected number of non-ZIKV microcephaly cases would be 0 . 7 cases [25] . In 2015 , the ZIKV epidemic year in the Northeast region , there were 1 , 142 registered microcephaly cases and a prevalence of 138 . 7 ( CI: 130 . 9–147 . 0 ) per 100 , 000 live births [42] , which represents a striking increase of the prevalence . Notably , we observed a high frequency of ZIKV positivity among DENV-immune mothers . This finding might reflect the high risk of exposure of this population to Aedes aegypti and , consequently , to arthropod-borne virus infection in general [26 , 27 , 40] . Additionally , the antibody-dependent enhancement ( ADE ) phenomenon may play a role in increasing the risk of ZIKV severe outcomes among individuals previously exposed to DENV [43–45] . The relationship between dengue and Zika cross-reactive antibody interactions may be complex: high dengue antibody titers might provide some protection against Zika infection , while low titers might favor enhancement . However , we acknowledge that case-control studies are not the proper epidemiological study design to accurately address this question . Ongoing prospective cohort studies may provide more conclusive responses in the near future . The main concern regarding serological assays is the potential for cross-reactivity of antibodies induced by DENV and ZIKV , posing a challenge in accurately differentiating infection by these viruses [11–18 , 20] . In our study , PCA using ZIKV and DENV serological assays as explanatory variables identified two main components: one strongly correlated with previous DENV infection and the other correlated with recent and previous ZIKV infection . Although most mothers of cases and controls included in our study had previous exposure to multiple flavivirus infections , the PCA results showed that the two factors are relatively uncorrelated , therefore representing different components of information . In addition , the scores for the component related to recent and past ZIKV infection were higher among mothers of cases than controls . Recently , there has been a growing body of work focusing on characterizing Zika and dengue immunological cross-reactivity on both binding and neutralization assays . Cross-neutralizing antibodies after a secondary flavivirus infection , whether dengue or Zika , have been observed by our group [40] and others [18 , 19] . Recent studies have demonstrated transient induction of flavivirus cross-neutralizing antibodies soon after recovery from a ZIKV or DENV infection in individuals who experienced secondary flavivirus infections [18 , 19] . However , little to no cross-neutralization has been detected in late convalescent samples ( >2 months ) [18] , confirming that ZIKV-specific NAbs develop after a ZIKV infection , even in the presence of pre-existing dengue exposure [17 , 18] . In a case-control study design , it is not possible to determine the time of maternal infection . Our finding of a high frequency of ZIKV markers ( as determined by PRNT ) and a low frequency of recent ZIKV infection suggest that there was a lag time between viral exposure and time of delivery . Hence , maternal samples collected at delivery potentially represent late convalescent serum , which probably explains the more specific neutralizing response for ZIKV observed in our study . The efficient maternal-fetal transfer of ZIKV and DENV NAbs observed by our group is consistent with previous published studies from the same and other settings [33 , 39] . This finding reflects the active transport of IgG across the placenta , a well-documented immune mechanism mediated by the neonatal Fc receptor ( FcRn ) , which is present on syncytiotrophoblast cells in the placental tissues [46] . Maternally acquired dengue-specific antibodies have been shown to play a dual role during infancy by first conferring protection at birth and then increasing the risk for severe dengue infection , as antibodies wane to subneutralizing levels [47] . Studies investigating the role of maternally transferred ZIKV antibodies in mediating protection or severe ZIKV or DENV disease during the first year of life are strongly needed . Additionally , data on the placental transfer ratio of dengue and Zika antibodies might be useful to infer the occurrence of congenital ZIKV infection in infants during the first year of life . In fact , the relative difference between maternal and infant ZIKV titers and maternal and infant DENV titers at later time points after birth ( ~6 months ) have been used as a marker of ZIKV infection of the fetus [41] . Infants born to dengue-immune mothers usually have a sharp decline in DENV antibody titers , which were acquired through maternal transfer , at early ages . However , if infected during pregnancy , ZIKV antibody titers in the newborns will remain high as the infant produces antibodies to the in-utero ZIKV infection [41] . Our study reported in-depth Zika and dengue serological profiles in a well-designed case-control study carried out in a hyperendemic area of dengue transmission . The overall strength of our approach is the well-characterized antibody responses for ZIKV and DENV1-4 as measured by PRNT , which is normally not feasible in studies analyzing large sample panels . In summary , we detected a strikingly high frequency of ZIKV exposure among mothers during the first wave of the Zika outbreak in Northeast Brazil . In addition , the majority of the mothers were immune to multiple flavivirus infections . This information provides insights regarding the immune status of the population in relation to a recent ZIKV introduction and more than three decades of DENV circulation . Additionally , PCA suggests relatively independency between the set of variables related to ZIKV and DENV infections , confirming minimal cross-reactivity antibody interactions in the PRNT assay . Considering the relatively low frequency of markers of recent ZIKV exposure at delivery , screening for ZIKV immune status should be performed in the early stage and throughout pregnancy to monitor congenital ZIKV syndrome in endemic areas . Innovative laboratorial diagnostic approaches for ZIKV and DENV infections are urgently needed for the guidance of clinical practice and public health purposes .
|
The epicenter of the ZIKV epidemic was located in Northeast Brazil ( 2015/16 ) and was followed by a space and time cluster of congenital microcephaly cases . This region is also a dengue hyperendemic setting . Laboratory confirmation of ZIKV infection during pregnancy is challenging due to cross-reactivity with other flaviviruses , especially dengue . The neutralization test , which is the gold standard to discriminate between these viruses , is time-consuming , performed in few laboratories and does not define the time when the infection occurred . This study described the serological markers of ZIKV and DENV among participants ( mothers and neonates ) of a microcephaly case-control study conducted in Northeast Brazil ( 2016 ) . Our results showed a strikingly high frequency of ZIKV exposure among mothers after the first wave of the ZIKV outbreak in this setting . Additionally , ZIKV and DENV immune status , as detected by the neutralization test , showed distinct patterns among pregnant women in this endemic area . We detected a low frequency of serological markers of recent ZIKV infection in samples collected just after delivery , highlighting the need for screening for ZIKV immune status in the early stage and throughout pregnancy to monitor congenital ZIKV syndrome .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"microcephaly",
"pathology",
"and",
"laboratory",
"medicine",
"enzyme-linked",
"immunoassays",
"pathogens",
"immunology",
"microbiology",
"viruses",
"research",
"design",
"developmental",
"biology",
"rna",
"viruses",
"antibodies",
"immunologic",
"techniques",
"morphogenesis",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"serology",
"proteins",
"medical",
"microbiology",
"birth",
"defects",
"case-control",
"studies",
"microbial",
"pathogens",
"immunoassays",
"congenital",
"disorders",
"biochemistry",
"flaviviruses",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"neonates",
"organisms",
"zika",
"virus"
] |
2019
|
Perinatal analyses of Zika- and dengue virus-specific neutralizing antibodies: A microcephaly case-control study in an area of high dengue endemicity in Brazil
|
Sporadic evidence suggests Notch is involved in cell adhesion . However , the underlying mechanism is unknown . Here I have investigated an epithelial remodeling process in the Drosophila eye in which two primary pigment cells ( PPCs ) with a characteristic ‘kidney’ shape enwrap and eventually isolate a group of cone cells from inter-ommatidial cells ( IOCs ) . This paper shows that in the developing Drosophila eye the ligand Delta was transcribed in cone cells and Notch was activated in the adjacent PPC precursors . In the absence of Notch , emerging PPCs failed to enwrap cone cells , and hibris ( hbs ) and sns , two genes coding for adhesion molecules of the Nephrin group that mediate preferential adhesion , were not transcribed in PPC precursors . Conversely , activation of Notch in single IOCs led to ectopic expression of hbs and sns . By contrast , in a single IOC that normally transcribes rst , a gene coding for an adhesion molecule of the Neph1 group that binds Hbs and Sns , activation of Notch led to a loss of rst transcription . In addition , in a Notch mutant where two emerging PPCs failed to enwrap cone cells , expression of hbs in PPC precursors restored the ability of these cells to surround cone cells . Further , expression of hbs or rst in a single rst- or hbs-expressing cell , respectively , led to removal of the counterpart from the membrane within the same cell through cis-interaction and forced expression of Rst in all hbs-expressing PPCs strongly disrupted the remodeling process . Finally , a loss of both hbs and sns in single PPC precursors led to constriction of the apical surface that compromised the ‘kidney’ shape of PPCs . Taken together , these results indicate that cone cells utilize Notch signaling to instruct neighboring PPC precursors to surround them and Notch controls the remodeling process by differentially regulating four adhesion genes .
Pattern formation in developing tissues requires cell signaling . A small number of signaling pathways are repeatedly utilized for cell fate decisions in developing tissues ( reviewed in [1] ) . In addition , cell signaling is also known to play a role in controlling cell sorting . For example , in the Drosophila wing , Hh signaling regulates cell segregation between anterior and posterior compartments ( reviewed in [2] ) , while Notch signaling is required for establishing a boundary that separates dorsal and ventral cells ( reviewed in [3] ) . In the Drosophila eye , Notch is required for a variety of developmental steps including rearranging pigment cells into hexagonal arrays [4] . All these observations raise the question of how Notch is involved in tissue remodeling . The observation that Notch is expressed in an epithelial sheet in the Drosophila embryo and continuously required for embryonic development after cell fate decision has led to speculation that Notch is involved in cell adhesion [5] . The behavior of primary pigment cells in the pupal eye also supports this view [4] . However , how Notch is involved in cell adhesion remains unclear . Evidence accumulated to date supports the notion that cell adhesion plays a direct role in tissue remodeling . As first noted by J . Holtfreter and later formulated in “Differential Adhesion Hypothesis” ( DAH ) by M . Steinberg: sorting behaviors of cells are driven by interfacial free energy arising from differential adhesion among cells [6] , [7] , [8] , [9] . In vivo observations support the DAH model . For example , in the Drosophila egg chamber , differential expression of E-cadherin determines localization of oocytes [10] , [11] . In the eye epithelium , homophilic interactions mediated by E- and N-cadherin direct a group of four cone cells to arrange in a pattern that minimizes surface free energy [12] . In the chick spinal cord , MN-cadherin is involved in sorting out motor neurons [13] . All these examples show that cadherins are directly responsible for cell sorting in a variety of tissues through homophilic interactions . On the other hand , more complex patterns involve more intricate mechanisms . For example , in the Drosophila pupal eye organizing pigment cells into hexagonal arrays requires two groups of heterophilic-interacting adhesion molecules: Hibris ( Hbs ) and Sticks-and-Stones ( Sns ) from the Nephrin group; Roughest ( Rst ) and Kin of Irre ( Kirre ) from the Neph1 group [14] . Nephrin and Neph1 are adhesion molecules of the IRM family within the immunoglobulin ( Ig ) superfamily and both proteins are essential for maintaining specialized junctions during kidney development in mammals [15] . Despite mounting evidence linking cell adhesion to cellular patterns , how cell-cell adhesion is regulated in developing tissues to generate a variety of cellular patterns remains unclear . This work describes a mechanism underlying an epithelial remodeling process in the Drosophila eye in which two primary pigment cells ( PPCs ) enwrap and isolate a group of cone cells from inter-ommatidial cells ( IOCs ) . This paper shows that Notch signaling controls transcription of two groups of adhesion genes in the Drosophila eye . Notch activates adhesion genes of the Nephrin group but suppresses those of the Neph1 group . Differential distribution of two groups of adhesion molecules is further facilitated by removal of one group of adhesion molecules by another group through cis-interactions , leading to complementary distribution of four adhesion molecules within two populations of cells . This work uncovers a link between cell signaling and tissue remodeling .
The Drosophila eye derives from an invaginated epithelium at the embryonic stage [16] . Photoreceptor neurons and lens-secreting cone cells are specified at late larval and early pupal stages . At 18 h after puparium formation ( APF ) , cone cells are surrounded by 4–5 inter-ommatidial cells ( IOCs ) , which have relaxed apical profiles ( Fig . 1A–A′ ) . Shortly , two cells adjacent to cone cells start to expand apical contacts with cone cells in most ommatidia . At 20 h APF , these two cells completely enwrap cone cells with a ‘kidney’ shape and they become two primary pigment cells ( PPCs ) ( Fig . 1B–B′ ) . As a result , cone cells are fully isolated from the rest of IOCs within the epithelial plane . Further rearrangement of IOCs gives rise to a one-cell wide hexagonal lattice of IOCs that fully separates ommatidia . Separation of ommatidia by IOCs will eventually serve to optically insulate the ommatidial array across the eye ( Fig . 1C–C′ ) . When Notch was depleted in all IOCs using RNAi , PPC precursors failed to enwrap cone cells . As a result , at 40 h APF , the cone cell cluster was found typically in direct contact with 4∼5 IOCs in an ommatidium ( Fig . 1D ) , indicating Notch is required for the assembly of ommatidia ( cone cells and PPCs ) . This phenotype is very reminiscent of the one seen in Nfa-g ( Fig . 1E ) . Nfa-g is a loss-of-function Notch allele in which the activity of Notch is retained throughout larval stages but lost within the pupal stage [17] . As a result , cone cells in the eye are not affected by the mutation . Previous studies indicate that in Nfa-g mutants two PPC precursors initially touch each other at both ends but they fail to establish contacts [4] , [18] , suggesting weakened adhesion between PPCs and/or adhesion between PPCs and cone cells . However , it has remained unclear how Notch is involved in cell-cell adhesion . The receptor Notch is broadly expressed in all cells in the early pupal eye [19] , [20] , [21] . In contrast , expression of the ligand Delta ( Dl ) is often cell type- specific and the protein is predominantly found within endocytic vesicles [21] . Consistent with the previous study [21] , Dl was detected in cone cells at 18 h APF ( Fig . 2A ) . Using a Dl-specific reporter , Dl transcript was detected in cone cells ( Fig . 2B ) . Especially , anterior cone cells had the highest level of Dl expression at this early stage ( Fig . 2B ) . By 24 h APF , although expression in the posterior cone cells was slightly increased , Dl expression in the anterior cone cells still remained the highest within the cone cell cluster ( Fig . 2C ) . To identify the cell types that receive active Notch signaling , a Notch activity reporter GBE-Su ( H ) m8-lacZ [22] was used . Consistent with expression of the Dl reporter , the Notch activity was detected in a significantly higher level in two cells adjacent to anterior-posterior cone cells than in other cells at 18 h APF ( Fig . 2D ) . These two cells were presumably the two PPC precursors . In particular , the highest Notch reporter activity was detected in the PPC precursor adjacent to the anterior cone cell within each ommatidium ( Fig . 2D ) . By 24 h APF , GBE-Su ( H ) m8-lacZ expression was found in both PPCs and the difference in the level of lacZ expression between these cells became less obvious than earlier stages ( Fig . 2E ) . Therefore , Dl transcription within the anterior and posterior cone cells is correlated with a high level of the Notch activity in the two PPC precursors . Previously it has been shown that hbs and sns , two genes from the Nephrin group , are transcribed in PPCs [14] , [23] . The pattern of the Notch activity is very reminiscent of hbs and sns expression . When an intracellular domain of Notch ( NICD , an activated form of Notch ) was expressed in a single PPC in the eye using a FLP-out technique [24] , the Rst protein level was increased 128% at the border between the target PPC and neighboring IOCs compared with wild type borders ( Fig . 3A and Table 1 ) , a phenotype very similar to over-expression of hbs in a PPC [14] . To test whether hbs transcription was activated upon activation of Notch , NICD was expressed in a single IOC . Upon activation of Notch , an ectopic activity of the hbs reporter P[w+]36 . 1 was observed in the target IOC ( Fig . 3B ) . These results indicate Notch is sufficient to activate hbs transcription . Consistently , when the Notch ligand Delta ( Dl ) was over-expressed in a single cone cell ( either anterior or posterior ) , the Rst level was increased about 71% at the border between the adjacent PPC and its neighboring IOCs ( Fig . 3C and Table 1 ) . When Dl was over-expressed in a single polar or equatorial cone cell , the Rst level was elevated about 81% at the two PPC-IOC borders encircling two PPCs ( Fig . 3D ) . These results indicate that the ligand Dl in cone cells is sufficient to activate hbs transcription in neighboring PPCs . To test the necessity of Notch in control of hbs transcription , Nfa-g mutant was used along with the hbs reporter P[w+]36 . 1 . In the wild type eye , the hbs reporter was detected in emerging PPCs as well as in cone cells [14] . In Nfa-g mutants , the hbs reporter activity was retained in cone cells but lost in PPC precursors at 40 h APF ( Fig . 3E ) . When GFP alone was expressed in single IOCs in the Nfa-g mutant , 12% of clones ( n = 86 , 4 eyes ) exhibited a kidney-shape seen in wild type PPCs . In contrast , when NICD ( activated Notch ) was expressed in single IOCs in the same mutant , 100% of clones ( n = 92 , 4 eyes ) exhibited kidney shape ( Fig . 3E″′ ) . In addition , these cells also expressed the hbs reporter ( Fig . 3E–E″ ) . These results indicate that Notch signaling is required for activation of hbs transcription . A similar effect was also observed with sns when Notch was activated in single IOCs in the Nfa-g mutant ( Fig . 3F–F″ ) . Taken together , these data indicate that Notch is both sufficient and necessary to activate transcription of both hbs and sns , the adhesion genes of the Nephrin group . When Notch was activated in a single IOC by expressing NICD , as expected , the Rst level was increased at IOC-IOC borders ( Fig . 4A–A″′ ) . Unexpectedly , the Rst level was reduced 40% at IOC-PPC borders ( Fig . 4A″′ and Table 1 ) . To test whether a reduction of the Rst protein seen at the PPC-IOC border is due to a reduction in rst transcription , a rst reporter ( rstF6-lacZ ) was used to monitor the rst activity . Upon activation of Notch in a single IOC , rstF6-lacZ was lost in the target cell ( Fig . 4B–B″′ ) , indicating Notch is sufficient to suppress rst transcription in IOCs . Since Notch is normally activated in PPC precursors , this result suggests that in the wild-type eye Notch suppresses rst in developing PPCs . Consistently , in Nfa-g mutants , rstF6-lacZ was expanded to all pigment cells surrounding cone cells ( Fig . 4C–C″ ) , indicating that Notch is necessary for suppressing rst in emerging PPCs . A similar effect was also seen with kirre when Notch activities were altered ( Fig . 4D–D″ and Table 1 ) . Taken together , these results indicate that Notch is both sufficient and necessary to suppress rst and kirre transcription in developing PPCs . It has been shown previously that genes coding for adhesion molecules of the IRM family are expressed in complementary cell types during cell rearrangement ( e . g . , 24 h APF ) : hbs and sns in PPCs; rst and kirre in IOCs [14] , [23] . The patterns of hbs and rst transcription at 18 h are similar to those at later stages ( e . g . , 27 h APF ) based on the hbs and rst reporters ( Fig . 5A–C ) . However , immune-staining using specific antibodies revealed striking differences in the distribution patterns of the Hbs and Rst proteins in the eye between 18 h and 40 h APF . Especially , both Hbs and Rst were present ubiquitously at a high level at all borders among epithelial cells at 18 h APF ( Fig . 5D–D″′ ) . This is in drastic contrast to later stages ( e . g . , 27–40 h APF ) when both Hbs and Rst were diminished at IOC-IOC and PPC-cone borders ( see below ) . At 20 h APF when two PPCs fully enwrapped the four cone cells , both Hbs and Rst remained at a high level at PPC-PPC and PPC-cone borders but slightly reduced at IOC-IOC borders ( Fig . 5E–E″′ ) . At 27 h APF , similar to earlier stages , both Hbs and Rst were enriched at PPC-IOC borders . In contrast , these proteins were reduced at PPC-PPC and PPC-cone borders and diminished at IOC-IOC borders ( Fig . 5F–F″′ ) . At 40 h APF , both Hbs and Rst proteins were again enriched at IOC-PPC borders but diminished at PPC-PPC and PPC-cone borders ( Fig . 5G–G″′ ) . They were undetectable at IOC-IOC borders ( Fig . 5G–G″′ ) . A similar dynamics in protein distribution was also observed with Sns and Kirre ( data not shown ) . These results indicate that four adhesion molecules are initially present in all epithelial cells at 18–20 h APF in the eye and removed from one group of cells at later stages . Therefore , distribution of Hbs , Sns , Rst and Kirre proteins undergoes a transition from ubiquitous to complementary distribution during epithelial remodeling . Hbs and Sns from the Nephrin group and Rst and Kirre from the Neph1 group co-localize at the border between PPCs and IOCs , and heterophilic interactions between these two groups of proteins in trans ( interactions between proteins from two adjacent cells or trans-interactions ) stabilize the adhesion complex on the membrane [14] , [23] . The observation that both Hbs and Rst were found in all IOCs at the beginning of cell rearrangement ( Fig . 5D–E″′ ) raises the question of how these IRM adhesion molecules interact with each other when placed in the same cell ( cis-interaction ) . To assess the effect of cis-interaction , Hbs was mis-expressed in the cells that normally express the counterparts Rst and Kirre . Upon expression of Hbs in a single IOC , the level of Rst was reduced about 63% at the PPC-IOC border and the number of vesicles was increased significantly in the target IOC ( Fig . 6A–A″ and Table 1 ) . Nevertheless , transcription of rst as assessed by the rst reporter rstF6-lacZ was not altered in the clone ( data not shown ) . Similarly , when Rst was mis-expressed in a single PPC that normally transcribes hbs and sns , the Hbs level on the membrane was reduced 93% and the number of vesicles increased markedly in the target PPC ( Fig . 6B–B″ and Table 1 ) . Similarly , the activity of the hbs reporter P[w+]36 . 1 was unchanged in the clone ( data not shown ) . These results suggest that , while heterophilic interactions between two groups of IRM adhesion molecules in trans stabilize both proteins on the membrane , interactions among these proteins in cis destabilize proteins on the membrane and promote turnover of these proteins . To assess the effect of cis-interactions on pattern formation , rst was mis-expressed in all PPCs using spa-Gal4 . Spa-Gal4 is known to drive expression of transgenes in cone cells and PPCs [25] . Upon expression of rst in cone cells and PPCs ( spa>rst ) , the hexagonal pattern of the eye was severely disrupted . While spatial organization of cone cells was mildly affected , various numbers of PPCs ( typically ranging from 1 to 3 ) were found adjacent to cone cells . More strikingly , IOCs failed to sort into single file . As a result , 2–3 rows of IOCs scattered in between ommatidia across the eye and the eye was extremely rough ( Fig . 6C ) . To exclude the possibility that the effect of over-expression was simply due to enhanced adhesion among cone cells and/or PPCs , N-cadherin was over-expressed in these cells using the same spa-Gal4 . N-cadherin is known to mediate adhesion among cone cells through homophilic interactions [12] . In contrast to Rst , over-expression of N-cadherin ( spa>N-cadherin ) only led to mild defects in IOCs and cone cells with largely intact PPCs ( Fig . 6D ) . To exclude the possibility that the severe defects seen in spa>rst are simply due to detrimental effects of the protein on the cells when expressed at a high level , Rst was over-expressed in IOCs using Gal4-54 ( 54>rst ) . In the 54>rst eye , IOCs were occasionally found in cluster with a bristle group . Nevertheless , the hexagonal pattern was only mildly affected ( Fig . 6E ) . Although we cannot exclude the possibility that different protein levels also contribute to the different phenotypes seen in these experiments , the results presented here strongly suggest that interference of IRM adhesion molecules by cis-interactions has a strong impact on the establishment of the hexagonal pattern . This work demonstrates that Notch controls transcription of IRM adhesion genes . On the other hand , Notch is also known to control transcription of multiple other genes during eye development . It is not clear whether loss of adhesion molecules is responsible for the PPC defects seen in Notch mutants . To address this issue , a rescue experiment was performed using Nfa-g as a background mutant and UAS-hbs as a rescue construct . When Hbs was expressed in a single cell adjacent to cone cells , 83% of the target cells ( n = 257 , 13 eyes ) elongated and the interface between the target cell and IOCs expanded in a manner similar to a wild type PPC ( Fig . 7A ) . Further , when Hbs was expressed in two cells adjacent to cone cells , in nearly all cases examined so far , Hbs positive cells fully enwrapped the cone cell group from the anterior and posterior sides resembling two wild type PPCs ( Fig . 7B ) . These results indicate that adhesion is sufficient to restore spatial relationship of cell clusters in Notch mutants . To test the necessity of cell-cell adhesion for formation of the spatial pattern of PPCs , hbs and sns double mutant was generated using snsZF1 . 4 and hbs459 mutant alleles ( see Materials and Methods ) . Large clonal patches generated using this double mutant together with ey-FLP led to extremely rough eyes in adults ( data not shown ) . Single PPCs mutant for both sns and hbs had a shorten PPC-IOC border and reduced apical surface ( Fig . 7C–C″′ ) . In addition , PPC-PPC border became curved . As a result , the apical profile of the target PPC became more rounded . These results indicate that sns and hbs are required for the normal ‘kidney’ shape of PPCs .
This work demonstrates that Notch is involved in cell-cell adhesion by regulating transcription of adhesion genes . Notch signaling is known to play a pleiotropic role in controlling cell fate during animal development [26] . The requirement of Notch during Drosophila embryonic development after cell fate decision has led to speculation that Notch is involved in cell adhesion [5] . This notion is supported by the behavior of PPCs in the pupal eye [4] . However , clear evidence linking Notch to cell adhesion has been lacking . This study shows that in the pupal eye Notch differentially controls transcription of four IRM adhesion genes . Notch activates transcription of hbs and sns but represses rst and kirre , leading to differential expression of IRM adhesion genes in two populations of cells: IOCs by default express rst and kirre; PPCs by activation of Notch signaling express hbs and sns . Heterophilic interactions between Hbs/Sns and Rst/Kirre proteins mediate preferential adhesion between IOCs and PPCs [14] , [23] . Therefore , Notch signaling sets up differential expression of adhesion genes ( Fig . 7D ) . This work also illustrates how a single signaling pathway transforms an initially homogeneous population of cells into two morphologically distinct groups of cells . In the wild-type eye , PPCs are polarized since PPCs without exception enwrap cone cells from anterior/posterior rather than from polar/equatorial sides . Data presented in this work suggest that asymmetric distribution of Dl in cone cells sets up PPC polarity . At the beginning of cell rearrangement ( ∼18 h APF ) , all IOCs that contact cone cells have access to Dl and express Hbs . However , asymmetric expression of Dl in cone cells creates a bias . IOCs that contact anterior-posterior cone cells receive a high level of Notch signaling ( thick red lines , Fig . 7D ) and produce more Hbs , which in turn boosts the ability of these cells to enwrap cone cells and gain more access to Notch signaling . In contrast , other IOCs that initially receive a low level of Notch signaling ( thin red lines , Fig . 7D ) are at a disadvantage and quickly lose competition to PPC precursors in enwrapping cone cells . As a result , Notch and Hbs create a positive feedback loop through which an initial small difference in Notch signaling is amplified , giving rise to PPCs exclusively enwrapping cone cells from anterior and posterior sides ( Fig . 7D ) . This work provides evidence that interactions between adhesion molecules from the Nephrin group and those from the Neph1 group in cis promote protein turnover . IRM adhesion molecules are known to form heterophilic interactions . Proteins from the Nephrin group bind in trans to proteins from the Neph1 group and trans-interactions among IRM adhesion molecules stabilize proteins on the membrane [14] , [23] . In contrast , cis-interactions among these proteins destabilize proteins on the membrane ( this work ) . Results presented herein support a model that cis-interactions provide a mechanism for removing counterpart proteins from the same cells ( Fig . 7D ) . After two PPC precursors completely surround the cone cell group ( e . g . , 20 h APF ) , these cells gain full access to Dl . In response to Notch signaling , PPC precursors constantly produce Hbs , which removes Rst from the same cells through cis-interaction . By the same mechanism , all other IOCs that are now denied access to Dl by default constantly produce Rst , which in turn clears Hbs from IOCs . Therefore , a combination of transcriptional regulation by Notch and post-translational mechanism by cis-interactions provides a mechanism for the transformation of initially ubiquitous distribution into complementary distribution of four adhesion molecules within two populations of cells ( Fig . 7D ) . cis-interactions observed in the Drosophila eye are very reminiscent of interactions between the Notch receptor and its ligand Dl . It has been shown that , in the Drosophila embryo and the eye imaginal disk , an increase of Dl in a Notch-expressing cell inhibits Notch signaling in a cell-autonomous fashion via cis-interaction [27] , [28] . In the Notch-Dl case , the level of Dl within a Notch-expressing cell determines the intensity of Notch signaling that cells receive , which in turn determines cell fates [27] , [28] . In the case of IRM adhesion molecules , the level of a protein from one group in a cell determines the amount of counterpart proteins from the other group on the membrane of the same cell , which alters cell-cell adhesion . More specifically , in the Drosophila eye cis-interactions remove remnant proteins and facilitate the differential distribution of IRM adhesion molecules without affecting cell fate . Despite different impact of cis-interactions on cell-cell interactions in both cases ( N-Dl versus IRM adhesion molecules ) , they share one common feature: presence of one protein interferes with the function of the counterpart protein in the same cell . What structural elements are involved in cis-interactions between these proteins and how cis-interactions lead to a reduction of protein activity still remain questions for further investigation . Although evidence presented in this work suggests a simple relationship among cell signaling , cell adhesion and cell shape , two observations highlight the complexity of PPC recruitment in the developing Drosophila eye . First , hbs can restore the ‘kidney’ shape of PPCs at a lower frequency than Notch ( or NICD ) in the Nfa-g mutant . This observation suggests that adhesion genes may not represent all the function of Notch in recruiting PPCs . Notch is known to have a wide range of target genes . In particular , several transcription factors are known targets of Notch for the determination of PPC cell fate [29] . Therefore , it is possible that additional effectors of Notch signaling are also involved in conferring on PPCs the ability to enwrap cone cells . Second , this work suggests a positive feedback loop that promotes selection of PPC precursors in the developing eye ( Fig . 7D ) . On the other hand , it has been shown recently that Hbs promotes Notch signaling by interacting with presenilin [30] . A potential more direct impact of Hbs on Notch signaling raises the possibility that there may exist a second positive feedback loop between Notch and Hbs: Notch activates hbs transcription and Hbs in return enhances Notch signaling , whereby initially a small difference of Notch signaling among IOCs is amplified , leading to separation of Hbs/Sns-expressing cells from those expressing Rst/Kirre . Whether the second positive feedback loop plays a role in PPC recruitment remains to be tested . This paper illustrates how a small number of cells utilize a single signaling pathway to instruct neighboring cells to surround them , whereby the centrally localized cells are isolated from other cells . Since isolation of a group of cells by another is commonly seen in developing tissues , a correlation between cell signaling and cell adhesion may be a more general mechanism for organizing cells during organ formation .
The sns and hbs double mutant snsZF1 . 4 hbs459 was generated for this work by recombining snsZF1 . 4 and hbs459 , a loss-of-function allele of sns and hbs , respectively , onto the second chromosome . Nfa-g , UAS-Notch RNAi , Dl-lacZ , y w hsFLP , UAS-nlsGFP and Act5C>y+>Gal4 UAS-GFP were provided by the Bloomington Stock Center . rst-Gal4 was obtained from National Institute of Genetics Fly Stock Center ( Japan ) . Other flies used: rstF6-lacZ [31] , spa-Gal4 [25] , UAS-N-cadherin [12] , snsZF1 . 4 and UAS-sns ( gift of Susan Abmayr ) , UAS-NICD ( gift of Cedric Wesley ) , UAS-NICD-lexA ( gift of Toby Lieber ) , P[w+]36 . 1 and hbs459 ( gift of Mary Baylies ) , UAS-hbs ( gift of Helen Sink ) , GBE-Su ( H ) m8-lacZ ( N-lacZ ) [22] , Gal-54 [23] , UAS-rst ( gift of Karl-F . Fischbach ) , UAS-kirre/duf ( gift of Marc Ruiz-Gomez ) , UAS-Dl ( gift of Marek Mlodzik ) and hsFLP MKRS ( gift of Matthew Freeman ) . Single cell clones for over-expressing a target gene were generated using a FLP-out technique as described previously [14] . To induce clones , pupae at 12 h APF were heat-shocked at 37°C in a water bath for 20 min . Clones were marked by GFP . The genotypes of clones are shown as follows: Loss-of-function clones were generated using a MARCM technique [32] . Clones were induced by heat-shocking third instar larvae at 37°C for 1 h . Clones were marked by GFP . The genotype of clones: yw hsFLP; FRT42D snsZF1 . 4 hbs459/FRT42D Gal80; tub-Gal4 UAS-mCD8-GFP/+ ( Fig . 7C ) . Immunostaining of the pupal eye was carried out as described [14] . Rat anti-Kirre ( 1∶5000 ) and rabbit anti-Hbs AS14 ( 1∶2500 ) were used as previously described [23] . Other primary antibodies: mouse anti-Rst Mab24A5 . 1 ( 1∶100 ) [33] , rabbit anti-Sns ( 1∶300 ) [34] and rabbit anti-lacZ ( 1∶2000; 5 Prime→3 Prime ) . Rat anti-DE-cadherin ( 1∶20 ) , mouse anti-Armadillo ( 1∶20 ) and mouse anti-Dl 9B ( 1∶20 ) were provided by Developmental Studies Hybridoma Bank at the University of Iowa . Secondary antibodies: Alexa 488 and Alexa 568 conjugated secondary antibodies ( 1∶5000; Molecular Probes ) ; Cy5 conjugated secondary antibodies ( 1∶1000; Jackson ImmunoResearch Laboratories ) . All images were captured using an Axioplan2 epi-fluorescence microscope equipped with an Axiocam digital camera ( Carl Zeiss , Inc . ) . Levels of membrane proteins were quantified using ImageJ [35] . Briefly , a long and narrow stripe that surrounded and closely followed the target border was carefully traced . The integrated density ( ID1 ) of the selected region was recorded using ImageJ . The background integrated density ( ID0 ) was recorded by moving the same selection box to a background region . The membrane protein level ( I ) reflected by integrated intensity per unit length was determined following I = 2* ( ID1−ID0 ) /L , where L is the perimeter of the selected region in pixel . For each experiment , the average intensity from 6 neighboring wild type borders was calculated and used as a control .
|
In developing tissues , one way to isolate a group of cells from the rest of the tissue is to induce a few neighboring cells to surround them . How centrally localized cells communicate with neighboring cells and how neighboring cells respond to signaling is not well understood . This work describes a mechanism underlying an epithelial remodeling process in the Drosophila eye in which two primary pigment cells ( PPCs ) with a characteristic ‘kidney’ shape enwrap and isolate a group of cone cells from inter-ommatidial cells ( IOCs ) . This paper shows that cone cells utilize Notch signaling to communicate with neighboring PPC precursors . In response to Notch signaling , PPC precursors activate transcription of hbs and sns , two genes coding for adhesion molecules of the Nephrin group that bind Rst and Kirre , adhesion molecules of the Neph1 group . At the same time , PPC precursors inactivate transcription of rst and kirre genes . In addition , binding of Hbs or Rst to its counterpart from the same cell ( cis-interaction ) destabilizes the protein complex and promotes removal of the counterparts from the membrane , leading to complementary distribution of four adhesion molecules within two populations of cells . Thus , Notch controls epithelial remodeling by differentially regulating four adhesion genes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"molecular",
"cell",
"biology",
"genetics",
"developmental",
"biology",
"biology"
] |
2014
|
Notch Controls Cell Adhesion in the Drosophila Eye
|
According to the Centers for Disease Control and Prevention ( CDC ) , one in twenty five hospital patients are infected with at least one healthcare acquired infection ( HAI ) on any given day . Early detection of possible HAI outbreaks help practitioners implement countermeasures before the infection spreads extensively . Here , we develop an efficient data and model driven method to detect outbreaks with high accuracy . We leverage mechanistic modeling of C . difficile infection , a major HAI disease , to simulate its spread in a hospital wing and design efficient near-optimal algorithms to select people and locations to monitor using an optimization formulation . Results show that our strategy detects up to 95% of “future” C . difficile outbreaks . We design our method by incorporating specific hospital practices ( like swabbing for infections ) as well . As a result , our method outperforms state-of-the-art algorithms for outbreak detection . Finally , a qualitative study of our result shows that the people and locations we select to monitor as sensors are intuitive and meaningful .
Since the time of Hippocrates , the “father of western medicine” , a central tenet of medical care has been to “do no harm . ” Unfortunately , the scourge of healthcare acquired infections ( HAI ) challenges the medical system to honor this tenet . When patients are hospitalized they are seeking care and healing , however , they are simultaneously being exposed to risky infections from others in the hospital , and in their weakened state are much more susceptible to these infections than they would be normally . Acquiring these infections increases the chances of either dying or becoming even sicker , which also lengthens the time the patient needs to stay in the hospital ( increasing costs ) . These infections can range from pneumonia and gastro-intestinal infections like Clostridium difficile to surgical site infections and catheter associated infections , which puts nearly any patient in the hospital at risk . Antibiotic treatments intended to aid in recovery from one infection , may open the door for increased risk of infection from another . Healthcare acquired infections are a significant problem in the United States and around the world . Some estimates put the annual cost between 28 and 45 billion US dollars per year in the US [1] . More importantly , they inflict a significant burden on human health . A recent study estimated more than 2 . 5 million new cases per year in Europe alone , inflicting a loss of just over 500 disability-adjusted life years ( DALYS ) per 100 , 000 population [2] . Given their burden and cost , their prevention is a high priority for infection control specialists . A simple approach to monitor HAI outbreaks would be to test every patient and staff in the hospital and swab every possible location for HAI infection . However , such a naive process is too expensive to implement . A better strategy is required to efficiently monitor HAI outbreaks . A recent review article [3] included 29 hospital outbreak detection algorithms described in the literature . They found these fall into five main categories: simple thresholds , statistical process control , scan statistics , traditional statistical models , and data mining methods . Comparing the performance of these methods is challenging given the myriad diseases , definitions of outbreaks , study environments , and ultimately the purpose of the studies themselves . However , the authors identify that few of these studies were able to leverage important covariates in their detection algorithms . For example , including the culture site or antibiotic resistance was shown to boost detectability . Past simulation based approaches [4] tackle optimal surveillance system design , by choosing clinics as sensors , to increase sensitivity and time to detection for outbreaks in a population . In contrast , our approach selects most vulnerable people and locations to infections as sensors to detect outbreaks in a hospital setting . Different kinds of mechanistic models have also been used for studying HAI spread [5 , 6 , 7 , 8] . Most of these are differential equation based models . We refer to [9] for a review of mechanistic models of HAI transmission . On a broader level , sensor selection problem for propagation ( of contents , disease , rumors and so on ) over networks has gained much attention in the data mining community . Traditional sensor selection approaches [10 , 11] typically select a set of nodes which require constant monitoring . Instead , in this paper , we select sensor set as well as the rate to monitor each sensor . Hence , our approach is novel from the data mining perspective as well . Recently Shao et al . [12] proposed selecting a set of users on social media to detect outbreaks in the general population . Similarly , Reis et al . [13] proposed an epidemiological network modeling approach for respiratory and gastrointestinal disease outbreaks . Other closely related data mining problems include selecting nodes for inhibiting epidemic outbreaks ( vaccination ) [14 , 15 , 16] and inferring missing infections in an epidemic outbreak [17] . We employ a simulation and data optimization based approach to design our algorithm and to provide robust bounds on its performance . Additionally , our simulation model is richly detailed in terms of the class of individuals and locations where sampling can occur . None of the prior works explicitly model the multiple pathways of infections for HAI outbreaks and fail in separating the location contamination and infections in people . We formalize the sensor set problem as an optimization problem over the space of rate vectors , which represent the rates at which to monitor each location and person . We consider two objectives , namely the probability of detection and the detection time , and show that the prior satisfies a mathematical property called submodularity , which enables efficient algorithms . In addition , we leverage data generated from a carefully calibrated simulation using real data collected from a local hospital . Our extensive experiments show that our approach outperforms the state-of-the-art general outbreak detection algorithm . We also show that our approach achieves the minimum outbreak detection time compared to other alternatives . To the best of our knowledge , we are the first to provide a principled data-driven optimization based approach for HAI outbreak detection . Though we validate our approach for a specific HAI , namely C . difficile , our general approach is applicable for other HAIs with similar disease model as well .
As previously mentioned , we propose a data-driven approach in selecting the sensors . There are multiple challenges in obtaining actual HAI spread data such as high cost , data sparsity , and the ability to safeguard patient personal information . For this reason , we rely on simulated HAI contagion data . We use a highly-detailed agent-based simulation that employs a mobility log obtained from local hospitals [18 , 19] to produce realistic contagion data . All the steps of this methodology are described in detail in [18] , and we summarize them below for completeness . Fig 1 shows a visualization of simulated HAI spread . In this simulation , people ( human agents ) move across various locations ( static agents ) as defined by the mobility log and spread HAI in stochastic manner . The simulation was developed in the following three steps: design of an in-silico or computer-based population and its activities , conceptualization of a disease model for a pathogen of interest , and the employment of a highly-detailed simulation . The following sections describe the data creation process in more detail . Recall that our goal is to select a set of agents as sensors , and the rate at which each such sensor should be monitored , such that future HAI outbreaks are detected with high probability , and as early as possible . However , these have to be selected within given resource constraints . We start with a formalization of these problems . Finding a minimum cost sensor set is a challenging optimization problem , and we present efficient algorithms by using the notion of submodularity . We first define some notations . Let bold letters represent vectors . Let P and L denote the sets of human agents and locations respectively; let n = |P ∪ L|—this will be the total number of agents in our simulations . Let B denote the budget on a number of samples that is permitted ( weighted by cost of agents ) , i . e , it is the sum of expected number of swabs to detect whether a location is contaminated or a human is infected . As mentioned earlier , the mobility logs are represented as a bipartite temporal network G ( P , L , E , T ) , with two partitions P and L representing agents , E representing who-visits-what-location relationship and T representing time/duration of the visit . We consider each agent to be a node in the temporal network G . Hence we use the terms node and agent interchangeably . Now , let c ∈ Rn , be the vector of costs , i . e . , c[v] is the cost of monitoring node v . Let r ∈ Rn be the vector of monitoring rates , where r[v] denotes that the probability that node v is monitored ( e . g . , swabbed ) each day . Finally , let Tmax denote the maximum time in each simulation instance . Unfortunately , Problems 1 and 2 are both computationally very challenging . In fact , both the problems can be proven to be NP-hard . Lemma 1 . Problem 1 is NP-hard . Lemma 2 . Problem 2 is NP-hard . We provide the proofs for both the lemmas in the supplementary , where we show that the NP-Complete SetCover problem can be viewed as a special case of both the Problems 1 and 2 . Since our problems are in the computational class NP-hard , they cannot be solved optimally in polynomial time even for simplistic instances , unless P = NP . The instances we need to consider are pretty large , so a naive exhaustive search for the optimal solution is also not feasible and will be too slow . Therefore , we focus on designing efficient near-optimal approximate solutions . We begin with Problem 1 . The function we are trying to optimize for problem 1 is defined over a discrete lattice , i . e . the rate vector r . Our approach is to show that this function is a submodular lattice function . The notion of submodularity , which is typically defined over set functions , can be extended to discrete lattice functions ( e . g . recenty in [32] ) . Informally , submodularity means that the objective value has a property of diminishing returns for a small increase in the rate in any dimension . It is important to note that submodularity for lattice functions is more nuanced than for simple set functions ( we define it formally in the Supplementary Information section ) . Fortunately , it turns out that this property implies that a natural greedy algorithm ( which maximizes the objective marginally at each step ) gurantees a ( 1 − 1/e ) -approximation to the optimal solution . Without such a property , it is not clear how to solve Problem 1 efficiently even for a small budget . We have the following lemma . Lemma 3 . The objective in Problem 1 is a submodular lattice function . The detailed description of the submodularity property and proof of lemma 3 are presented in the supplementary . Our HaiDetect algorithm for Problem 1 selects the sensors to be monitored and rates such that nodes which tend to get infected across multiple simulation instances have higher infection rates . Specifically , at each step , HaiDetect selects the node v and the rate r among all possible candidate pairs of nodes and rates , such that the average marginal gain is maximized . HaiDetect keeps adding nodes and/or increasing the rates to monitor the selected nodes until the weighted sum of the rates is equal to the budget B . The detailed pseudocode is presented in Algorithm 1 . Algorithm 1 HaiDetect Require: I , budget B 1: for each feasible initial vector r0 do 2: Initialize the rate vector r = r0 3: while ∑v r[v] ⋅ c[v] < B do 4: Find a node v and rate r maximizing average marginal gain 5: Let r[v] = r 6: Remove all candidate pairs of nodes and rates which are not feasible 7: Return the best rate vector r HaiDetect has desirable properties in terms of both effectiveness and speed . The performance guarantee of HaiDetect is given by the following lemma . Lemma 4 . HaiDetect gives a ( 1-1/e ) approximation to the optimal solution . The lemma above gives an offline bound on the performance of HaiDetect , i . e . , we can state that the ( 1-1/e ) approximation holds even before the computation starts . We can actually obtain a tighter bound by computing an empirical online bound ( once the solution is obtained ) which can be derived using the submodularity and monotonicity of Problem 1 . For us to state the empirical bound , let us define some notations . Let the solution selected by HaiDetect for a budget B be r ^ . Similarly , let the optimal vector for the same budget be r* . For simplicity , let the objective function in Problem 1 be R ( ⋅ ) . For all nodes v and for a ∈ [0 , 1] , let us define Δv as follows: Δ v = max a[ R ( r ^ ∨ a · χ { v } ) - R ( r ^ ) ] ( 8 ) Similarly let us define σv as the argument which maximizes Δv σ v = arg max a[ R ( r ^ ∨ a · χ { v } ) - R ( r ^ ) ] ( 9 ) Now , let δ v = Δ v c [ v ] · σ v . Note that for each node v , there is a single δ . Let the sequence of nodes s1 , s2 , … , sn be ordered in decreasing order of δv . Now let K be the index such that θ = ∑ i = 1 K - 1 c [ s i ] σ s i ≤ B and ∑ i = 1 K c [ s i ] σ s i > B . Now the following lemma can be stated . Lemma 5 . The online bound on R ( r* ) in terms of the current rate r ^ assigned by HaiDetect is as follows: R ( r * ) ≤ R ( r ^ ) + ∑ i = 1 K - 1 Δ s i + B - θ c [ s K ] σ s K Δ s K The lemma above allows us to compute how far the solution given by HaiDetect is from the optimal . We compute this bound and explore the results in detail in the Results section . In addition to the performance guarantee , HaiDetect’s running time complexity is as follows . Lemma 6 . The running time complexity of HaiDetect is O ( c ⋅ B2 ( |P| + |L| ) ) , where c is the number of unique initial vectors r0 , B is the budge , P is the set of human agents and L is the set of locations . Note that the constant c is much smaller than the total population , i . e . , c << |P| + |L| in our case as infections are sparse and we do not need to consider agents and locations which never get infected . The most expensive computational step in Algorithm 1 is the estimation of the node v and rate r that gives the maximum average marginal gain ( Step ( i ) of 1 ( b ) ) . This can be expedited using lazy evaluations and memoization . Hence , the algorithm is also quite fast in practice . Moreover , it also embarrassingly parallelizable . The steps ( a ) and ( b ) for each initial vector can be performed in parallel . We also propose a similar algorithm HaiEarlyDetect for Problem 2 . The main idea here is that we assign higher rates to nodes which tend to get infected earlier in many simulation instances . The pseudocode for HaiEarlyDetect is presented in Algorithm 2 . Algorithm 2 HaiEarlyDetect Require: I , budget B 1: for each feasible initial vector r0 do 2: Initialize the rate vector r = r0 3: while ∑v r[v] ⋅ c[v] < B do 4: Find a node v and rate r minimizing the average detection time 5: Let r[v] = r 6: Remove all candidate pairs of nodes and rates which are not feasible 7: Return the best rate vector r As shown in Algorithm 2 , HaiEarlyDetect optimizes the marginal gain in the objective in Problem 2 in each iteration . It turns out that the objective in Problem 2 is not submodular . However , as shown by our empirical results , the greedy approach we propose works very well in practice and outperforms the baselines . Moreover , it too runs fast in practice as the same optimization techniques discussed earlier for HaiDetect applies to HaiEarlyDetect as well .
In the previous section , we discussed two types of bounds on the performance of HaiDetect . Here we show how far the solution given by HaiDetect is from the optimal value for various budgets . For this experiment , we ran HaiDetect on a set of 100 simulations and computed the value of the objective in Problem 1 for the resulting rate vector . We also computed the overall bound , based on ( 1 − 1/e ) approximation and the empirical bound as per Lemma 5 . Since the objective value cannot exceed the number of simulations , we also compute the lowest bound as the minimum of two bounds and the number of simulations . We repeat the experiment for budget size from 1 to 50 . The resulting plot is presented in Fig 5 . Fig 5 highlights several interesting aspects . First of all , we can see that the online bound is always tighter than the offline bound . Moreover , we also observe that as the performance of HaiDetect reaches close to the optimal ( with increase in budget ) , the online bound becomes more and more tight until both the performance and bound are equal , indicating the values of the budget for which HaiDetect solves Problem 1 optimally . This results demonstrates that HaiDetect can accurately find sensors which can detect any observed outbreaks given sufficient budget . Given that HaiDetect is near-optimal for the observed outbreaks , we evaluate its effectiveness in detecting unseen ( “future” ) outbreaks . Here , we compare the performance of HaiDetect and Celf with respect to the budget on “unobserved” simulations . For this experiment , we performed a 5-fold cross validation on 200 simulations . Specifically , we divided the simulations into 5 groups , and at each turn we selected the sensors in the first four groups and computed the sum of outbreak detection probability as shown in Eq 1 in the fifth group ( the test set ) . Then we normalize the resulting sum of outbreak detection probability by total number of simulation instances in the the same group . The normalized value can be intuitively described as the average probability of detecting a future outbreak . We repeat this process five times ensuring each group is used for success evaluation . We then compute the overall average and its standard error . We repeat the entire process for the budgets from 1 to 50 . The result of our experiment is show in Fig 6 . The first observation is that HaiDetect consistently outperforms Celf for all values of the budget . The disparity between the methods is more apparent for larger values of budget . The difference in quality of the sensors can be explained by the fact that Celf only assigns rate of 0 or 1 . However , HaiDetect can strategically assign non-integer rates so as to maximize the likelihood of detection . We can also observe that the standard error for the HaiDetect decreases and is negligible for larger budgets . However , it is not the case for Celf . This shows that not only the quality of sensors detected by HaiDetect is better , but it is more stable as well . Finally , we see that probability of an outbreak being detected by sensors selected by HaiDetect is 0 . 96 when budget is equal to 50 , whereas it is only around 0 . 75 for Celf . Similarly , a budget of only 25 is required to detect an outbreak with probability of 0 . 8 for HaiDetect . For the same budget , sensors selected by Celf detect cascades with probability of 0 . 55 . The result highlights that HaiDetect produces more reliable monitoring strategy for HAI outbreak detection . Here , we investigate the change in performance of HaiDetect and Celf as the number of simulations used to detect the sensor increases . For this experiment , we used 150 distinct simulations . We divided the simulations into two categories , training and testing sets . We used the cascades in the training set to select the sensors and used the ones in the testing set to measure quality . First we decided on a budget of 10 and training size of 10 cascades . We ran both HaiDetect and Celf for this setting and measured the quality using the cascades in the testing set . We then increased the training size by 10 till we reached the size of 100 . We repeated the same procedure for budgets of 30 and 50 . We compute the average probability of detection in the same manner as described above . Fig 7 summarizes the result . We can observe that HaiDetect outperforms Celf consistently . It reinforces the previous observation that HaiDetect selects good sensors for the HAI outbreak detection . An interesting observation is that the performance tails off after training size of 20 for larger budgets , which implies that not many cascades have to be observed before we can select good quality sensors . This is an encouraging finding as gathering large number of real cascades of HAI spread is not feasible . Next we study the change in performance of HaiDetect with the training size for various budget sizes . Here we tracked the performance of HaiDetect for budgets of 10 , 30 , and 50 for training sets of various size . The result is summarized in Fig 8 . As shown in the figure , the difference between peformance of HaiDetect for budgets 30 and 10 is much larger than that for budgets 50 and 30 . The normalized objective , or the probability of detection , is close to 1 at budget 50 , indicating that monitoring sensors at rates assigned by HaiDetect detects almost all the HAI outbreaks . Hence , in expectation , roughly 50 swabs a day is enough to monitor an outbreak in a hospital wing . Again , we observe that performance of HaiDetect tails off after the training size of 20 . It provides extra validation for the observation that a limited number of observed cascades are enough to select high quality sensors . A desirable property of sensors is that they aid in early detection of outbreaks . Here we study the average detection time of future outbreaks using the sensors and rates selected by HaiEarlyDetect . In this experiment , we first divided our simulations into equally sized training and testing sets , each having 100 simulations . We ran HaiEarlyDetect on the training set to detect sensors and rates at which to monitor them . Then , we monitored the selected sensors at the inferred rates and measured the detection time for each simulation in the testing set . We repeated the entire process for various budgets . The detection time averaged over 100 simulated outbreaks in the testing set is summarized in Fig 9 and the variance in the detection time is shown in Fig 10 . As shown in the figure , as the budget increases the average detection time decreases . According to our results , the average time to detect an outbreak in the testing set while monitoring sensors selected for budget of 1000 is roughly six days . This is impressive considering the fact that monitoring all agents results in detection time of 4 days monitoring all of more than 1200 nurses results in detection time of 8 days . Hence , monitoring these sensors detect the outbreak earlier with fewer budget . Another advantage of our sensors is that they are diverse . Significant proportion of the selected sensors include patients and fomites , which are easier to monitor than the nurses . Hence , monitoring the sensors selected by HaiEarlyDetect also has an economic advantage . An interesting observation seen in Fig 10 is that the variablity in average detection time decreases with the increase in budget . Hence , we expect the performance of our sensors to be fairly consistent in detecting future outbreaks for larger budgets . Moreover , the median time to detect an outbreak ( as shown by the box plots ) is always less than the average . Hence , we expect that performance of HaiEarlyDetect to be generally better than that suggested by the average detection time . For budget of 1000 , the median detection time is just 5 days . Note that monitoring all agents results in detection time of 4 days . This implies that in practice our approach requires only 1000 swabs per day to detect an outbreak within a single day of the first infection . An interesting question is how many potential cases can be prevented by monitoring the sensors selected by HaiEarlyDetect . Here we study how many nodes get infected before an outbreak is detected and how many potential infections can be prevented by monitoring our sensors for various budgets . As in the previous experiment , for a given budget , we leverage 100 simulations to select sensors and their monitoring rates . Once the sensors are selected , we count the number of infections that occur in a test simulation before a sensor is infected and how many further infections occur following the infections of sensors . We then average these numbers over 100 test simulations . The results are summarized in Table 2 . As shown in Table 2 , for the budget of 10 samples/swabs , 4 . 31 potential future infections could prevented . Note that there are only 23 infections on average per simulation . For the budget of only 200 , 15 . 02 infections could be prevented , which is about 66% of potential infections . The number goes up to 17 , or 74% for the budget of 1000 . The result shows that even for a low budget ( less than 200 swabs per day ) , our approach could help prevent a significant number of future infections . Next we study the types of agents that are selected by HaiDetect as sensors . For this experiment , we use 100 randomly selected simulations to detect sensors for a wide range of budgets . After the sensors are selected , we sum up the rates of each category of agents like nurses , doctors , patients , and so on . Fig 11 ( a ) shows the distribution of sensor allocation for each category of agents at low budgets . We observe that for a budget of 10 , nearly 60% of the total budget is spent on selecting nurses . Since nurses are the most mobile agents , the result highlights the fact that HaiDetect selects the most important agents as sensors early on . Similarly , Fig 11 ( b ) shows the distribution of sensors for higher budgets . Here we observe that nearly 35% of the budget is allocated for nurses . Fomites and patients have roughly equal allocations of about 20% . 17% of the budget is allocated to doctors . The rest of the categories have minimal allocation . The distribution shows that HaiDetect selects heterogeneous sensors including both people and objects/locations as intended . Finally , we are also interested on the scheduling implications of the sensors selected by HaiDetect . To this end , we measure the aggregated proportion of budget assigned to each rate for the sensors we select . The results are summarized in Fig 12 . As shown in Fig 12 ( a ) , most of the sensors have rate of 0 . 1 . Very few sensors have rate from 0 . 2 to 0 . 5 . Finally , there is a sudden spike at rate = 1 . 0 . When we look at rate distribution for each category separately , interestingly we observe that only nurses have rates of 1 . 0 . This implies that certain nurses have to be monitored each day to detect HAI outbreak . The reason behind this unexpected behaviour can be attributed to the fact that the hospital from where the mobility log was collected , required all the nurses to attend a daily meeting . Hence , all the nurses were in contact with each other every day and it is likely that nurses infect each other in case of an outbreak . Hence , there is an advantage in monitoring some of the nurses everyday to quickly detect HAI outbreak .
Effective and early detection of HAI outbreaks are important problems in hospital infection control , and have not been studied systematically so far . While these are challenging problems , understanding their structure can help in designing effective algorithms and optimizing resources . Current practices in hospitals are fairly simple , and do not attempt to optimize resources . Our algorithms perform better than many natural heuristics , and our results show that a combination of data and model driven approach is effective in detecting HAIs . Since there is limited data on disease incidence , good models and simulations play an important role in designing algorithms and evaluating them .
|
Healthcare acquired infections ( HAIs ) lead to significant losses of lives and result in heavy economic burden on healthcare providers worldwide . Timely detection of HAI outbreaks will have a significant impact on the health infrastructure . Here , we propose an efficient and effective approach to detect HAI outbreaks by strategically monitoring selected people and locations ( sensors ) . Our approach leverages outbreak data generated by calibrated mechanistic simulation of C . difficile spread in a hospital wing and a careful computational formulation to determine the people and locations to monitor . Results show that our approach is effective in detecting outbreaks .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"gut",
"bacteria",
"medical",
"personnel",
"sociology",
"social",
"sciences",
"health",
"care",
"simulation",
"and",
"modeling",
"health",
"care",
"providers",
"systems",
"science",
"mathematics",
"network",
"analysis",
"social",
"networks",
"nosocomial",
"infections",
"bacteria",
"allied",
"health",
"care",
"professionals",
"research",
"and",
"analysis",
"methods",
"clostridium",
"difficile",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"epidemiology",
"agent-based",
"modeling",
"people",
"and",
"places",
"professions",
"nurses",
"population",
"groupings",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"organisms"
] |
2019
|
Fast and near-optimal monitoring for healthcare acquired infection outbreaks
|
The aetiological diagnostic of fevers in Laos remains difficult due to limited laboratory diagnostic facilities . However , it has recently become apparent that both scrub and murine typhus are common causes of previous undiagnosed fever . Epidemiological data suggests that scrub typhus would be more common in rural areas and murine typhus in urban areas , but there is very little recent information on factors involved in scrub and murine typhus transmission , especially where they are sympatric - as is the case in Vientiane , the capital of the Lao PDR . We therefore determined the frequency of IgG seropositivity against scrub typhus ( Orientia tsutsugamushi ) and murine typhus ( Rickettsia typhi ) , as indices of prior exposure to these pathogens , in randomly selected adults in urban and peri-urban Vientiane City ( n = 2 , 002 , ≥35 years ) . Anti-scrub and murine typhus IgG were detected by ELISA assays using filter paper elutes . We validated the accuracy of ELISA of these elutes against ELISA using serum samples . The overall prevalence of scrub and murine typhus IgG antibodies was 20 . 3% and 20 . 6% , respectively . Scrub typhus seropositivity was significantly higher among adults living in the periphery ( 28 . 4% ) than in the central zone ( 13 . 1% ) of Vientiane . In contrast , seroprevalence of murine typhus IgG antibodies was significantly higher in the central zone ( 30 . 8% ) as compared to the periphery ( 14 . 4% ) . In multivariate analysis , adults with a longer residence in Vientiane were at significant greater risk of past infection with murine typhus and at lower risk for scrub typhus . Those with no education , living on low incomes , living on plots of land with poor sanitary conditions , living in large households , and farmers were at higher risk of scrub typhus and those living in neighborhoods with high building density and close to markets were at greater risk for murine typhus and at lower risk of scrub typhus past infection . This study underscores the intense circulation of both scrub and murine typhus in Vientiane city and underlines difference in spatial distribution and risk factors involved in the transmission of these diseases .
Scrub typhus and murine typhus are important but under-recognized treatable causes of fever , morbidity and mortality in South-East Asia [1] , [2] , [3] , [4] . There has been a recent resurgence of interest in these diseases , which both cause undifferentiated fever , headache and myalgia progressing , in a minority , to jaundice , pneumonitis and meningo-encephalitis [5] , [6] , [7] , [8] , [9] . Scrub typhus , caused by Orientia tsutsugamushi , occurs in Asia and northern Australia and is transmitted by the bites of infected trombiculid mites [10] . Murine typhus , caused by Rickettsia typhi , occurs globally and is transmitted through the infected flea bite site or by scratching infected faeces into the skin [11] , [12] . Scrub typhus and murine typhus were first differentiated , in Malaysia , in 1936 [13] . Chiggers and rodents are thought to be most important reservoirs of scrub typhus and murine typhus infection , respectively . Studies suggest that scrub typhus is more common in rural areas and murine typhus in urban areas [11] , [13] , [14] , [15] , [16] , [17] , [18] , [19] but there is very little recent information on the epidemiology of scrub and murine typhus in places where both diseases occur . The Lao People's Democratic Republic ( Laos ) is situated mostly east of the Mekong River and borders Thailand , Cambodia , Burma , China and Vietnam . The majority of the population ( 88% ) of 5 . 6 million people lives in rural areas ( 2005 census from National Statistics Centre ) . Vientiane , the capital of Lao PDR is the most populated urban area in the country , with less than 300 , 000 inhabitants . The diagnosis of non-malarial fevers in Laos remains difficult due to limited laboratory diagnostic facilities . In 2000 , the main differential diagnoses for adults admitted with fever to hospital were slide-positive malaria or slide-negative ‘malaria syndrome’ and , in both situations patients were treated with antimalarials , with additional antibiotics for those with ‘malaria syndrome’ . Since then it has become apparent that both scrub and murine typhus are common causes of fever in Laos [20] , [21] , [22] . In Mahosot Hospital , Vientiane , among 427 adults admitted with unexplained fever , 14 . 8% and 9 . 6% had serological evidence for scrub typhus and murine typhus , respectively [20] . As these diseases are usually relatively straightforward and inexpensive to treat with short courses of doxycyline , their recognition in Laos raises the prospect that a significant proportion of non-malarial fevers can be diagnosed and treated relatively inexpensively . However , there is very little recent information on the epidemiology of scrub typhus and murine typhus especially where they are sympatric , as is the case in Vientiane , and there is a need for greater understanding of contrasting risk factors involved in scrub and murine typhus transmission . Therefore , we analyzed serological data from a randomly selected population of adults living in different neighbourhoods in Vientiane to determine the frequency of IgG seropositivity against scrub and murine typhus as indices of prior exposure to these pathogens .
Vientiane Capital City ( VCC ) refers to the province that includes the Vientiane urban agglomeration , as well as surrounding smaller urban areas and rural villages [25] . Urban Vientiane only was estimated to have 277 , 000 inhabitants in 2005 . This urban area is composed of 148 villages ( ‘ban’ in Lao ) , which are the primary administrative units and constituted the primary sampling unit of this seroprevalence survey . To define level of urbanization of neighbourhood in Vientiane city , we used some indicators based on 1995 and 2005 census and a variety of GIS-based indicators derived from 1999 aerial photographic coverage [26] rather than a single common indicator like population density . We selected thirteen indicators: built-up density , density of population , changes in built-on surface area between 1981 and 1999 , proportion of public infrastructure buildings , proportion of trade buildings , number of markets in proximity , distance to the city centre via the road network , average distance of every building to the road network , access to running water , electricity and toilets , proportion of concrete houses , and proportion of the population involved with agricultural activities . Using a prior hierarchical classification , three categories of neighborhoods in Vientiane were identified: 1 ) the central zone; 2 ) the first urbanized belt; and 3 ) the second urbanized belt , with respectively 25 , 67 and 56 neighbourhoods in each area . The first urbanized belt clearly differed from the central zone by a smaller proportion of public infrastructures , trade buildings and concrete houses . The first urbanized belt differed from the second belt by a higher density of built-up and of population and by household facilities ( such as running water , electricity and modern toilettes ) of much better quality . To carry out the seroprevalence survey , nine neighborhoods were selected in each urban category ( i . e . a total of 27 neighborhoods ) as representative of the variability of the urban population . In each neighborhood , households were selected randomly from a list of households . Within each selected household , one adult ( ≥35 years ) was randomly selected . To measure the association of place of residence with typhus seroprevalence rates , study participants were limited to adults who claimed continued residence in a single village of the study area for a minimum of five years . The population survey techniques have been described elsewhere [23] , [24] . Ethical approval for the study was granted by the Lao National Ethics Committee for Health Research in Lao PDR ( No 046 ) and the Oxford University Tropical Research Ethics Committee ( OXTREC 003-06 ) . All participants gave informed written consent prior to survey administration and sample collection . The survey took place in February and March 2006 . Standardized questionnaires were administered to assess demographic and socioeconomic information from the study population . Individual information about sex , age , origin , education level , occupation , length of residence in Vientiane city , contact with rats were collected . Information about household size and the presence of rats around the house and the sanitary condition of the household plot ( presence of rubbish and animal excrement ) were also recorded . A household deprivation index was developed from household characteristics ( e . g . house building materials , access to running water , types of cooking energy , possession of motorbike , car , refrigerator , washing machine and computer ) . Using Multiple Correspondence Analysis followed by Hierarchical Ascendant Classification , households incomes were classified in three categories: household with poor income ( 11% of sampled population ) , with intermediate income ( 61% ) and those with high income ( 28% ) . The distance from the house to the closest market was measured in a Geographical Information System and the density of buildings in residential neighborhood was also calculated from data derived from 1999 aerial photographic coverage [26] . This density - which ranged from 5 to 100% - was grouped into tertiles ( with thresholds of 65% and 84% ) and classified as low , intermediate and high density . For every surveyed adult , blood sample of approximately 75 µl were collected by fingerpick , absorbed on Proteinsaver™ filter papers ( Whatman plc , Maidstone , UK ) and stored at −80°C until use . Two discs of six mm diameter were cut , using a hole punch , from the centre of the dried filter paper blood spot and eluted at 37°C overnight in 500 µl of phosphate buffered saline ( PBS ) corresponding to a 1/25 dilution of original serum . The Rickettsia Scrub Typhus Group IgG ELISA ( E-RST01G , PanBio Diagnostics , Brisbane , Australia ) was used for the detection of anti-O . tsutsugamushi IgG antibodies and the manufacturers instructions were followed [27] , [28] . Anti-R . typhi antibody detection used an in house typhus group IgG ELISA technique [17] . In brief , one half of each 96 well microtiter plate was coated with 100 µl/well of R . typhi ( Wilmington ) whole cell antigen ( 1∶3000 dilution ) and 100 µl PBS was added per well to those in the other half of the plate . The plates were covered with a plastic lid and stored at +4−8°C for 2 days , washed 3 times with wash buffer ( 0 . 1% Tween 20 in PBS ) , blocked with 5% skim milk ( Cadbury , Bournville , Worcs . , UK ) in wash buffer ( dilution buffer ) and incubated at 37°C for one hour . Filter paper elutes were diluted with dilution buffer to a working concentration of 1∶100 , transferred to the plates and incubated at room temperature for 1 hour followed by 5 washes with wash buffer . The wells were incubated with an HRP _abelled affinity-purified antibody to human IgG ( H+L ) ( KPL , Maryland , USA ) at a dilution of 1∶2000 for one hour at room temperature . After washing 5 times , 100 µl/well of a peroxidase substrate , 2 , 2-azino-di-[ethylbenzthiazoline sulfonate] ( ABTS ) ( KPL ) was added and the plate incubated in the dark for 30 minutes at room temperature . 100 µl/well of ABTS stop solution ( KPL ) was added and the plate read immediately using a Multiskan EX ELISA reader ( Labsystems , MA , USA ) at 405 nm . The same ELISA plate reader was used to measure absorbance for scrub typhus group IgG assays . Equivocal results in both tests were repeated once . If the repeat test result remained as equivocal it was considered as negative in the statistical analysis . To determine the concordance of the two ELISA techniques using sera and filterpaper bloodspot elutes , both ELISA techniques were performed on these samples collected as a part of the study of Phetsouvanh et al . [21] from the same patients at the same time point . Proportions and 95% Confidence Intervals ( CI ) were calculated , taking into account the two-stage sample design . Differences in seropositivity between areas were calculated by the Pearson chi-squared test . Potential factors associated with scrub typhus and murine typhus past exposure were explored first using a bivariate analysis and , secondly , using multivariate logistic regression . Multivariate logistic regressions were performed using Intercooled Stata 10 ( Stata Corporation , College Station , Tx , USA ) with fitting of a random-effect logit model at the neighborhood scale . Odds Ratios ( OR ) and 95% CI were calculated . Only finals models of multivariate logistic regression with significant risk factors ( and with adjustment on age and sex ) were presented in this paper . A p value of <0 . 05 was considered as significant . Maps were generated using Geographic Information System ( GIS ) ArcGis 9 . 2 software ( ESRI , USA ) .
A sample of 2 , 002 adults was included in the study with a mean age ( range ) of 50 . 6 ( 35–90 ) years . Only 9% of population was excluded from the surveyed sample because of a length of residence less than five years . The population was older in the central zone than in the first or second urbanized belt with 18% , 16% and 13% of the population aged >65 years , respectively ( Table 1 ) . The sex ratios did not statistically differ ( p = 0 . 26 ) within the city , although the proportion of women was slightly higher in the central zone ( 63% ) compared to the rest of the city ( 58% ) . The proportion of non-Lao people was slightly - but not statistically ( p = 0 . 08 ) - higher ( 8% ) in the central zone compared to the rest of the city ( 5% ) . The population living in the second urbanized belt had a significant lower education level than those living in the central zone or first urbanized belt since 30% versus 40% , respectively , attended secondary school . Income of sampled households varied by the extent of urbanization: households with high income were much more frequent in the central zone and in the first urbanized belt ( 33% and 31% respectively ) than in the second urbanized belt ( 18% ) . Forty-three percent of sampled adults had lived in Vientiane for more than two thirds of their lifetime , without significant variation ( p = 0 . 09 ) across the city . Comparison of anti-scrub typhus IgG ELISA assays using 47 sera and filterpaper bloodspot pairs demonstrated agreement ( i . e . positive or negative for IgG against O . tsutsugamushi ) for 45 ( 96% ) – one pair was negative for IgG from filterpaper but positive from serum and one pair positive for IgG from filterpaper but negative from serum . Comparison for anti-murine typhus IgG ELISA assays using 45 sera and filterpaper bloodspot pairs demonstrated agreement for 42 ( 93% ) – two pairs were negative for IgG from filterpaper but positive from serum and one pair positive for IgG from filterpaper but negative from serum . Therefore , these anti-typhus IgG ELISAs using filterpaper elutes gave good agreement with results obtaining using sera . The overall percentage of scrub typhus IgG antibodies was 20 . 3% ( CI = 18 . 1–22 . 5 ) and 20 . 6% ( CI = 17 . 4–23 . 8 ) for murine typhus IgG antibodies ( Table 2 ) . Four percent of samples had IgG antibodies against both scrub and murine typhus . The prevalence of scrub typhus IgG antibodies was significantly higher ( p<0 . 01 ) among people living in the periphery than in the central zone: 13 . 1% positive in the central zone as compared to 16 . 8% and 28 . 4% for first and second urbanized belts , respectively ( Table 3 , Figure 1 ) . In contrast , seroprevalence of murine typhus IgG antibodies was significantly higher ( p<0 . 01 ) in the central zone ( 30 . 8% ) as compared to the first ( 20 . 1% ) and second ( 14 . 4% ) urbanized belts ( Table 3 , Figure 2 ) . Two villages ( Bonangua and Somvang Tay ) located in the extreme north and extreme southeast of the city had very high seroprevalence of scrub typhus , exceeding 35% ( Figure 1 ) . Highest seroprevalences ( >32% ) against murine typhus occurred in four villages ( Anou , Thongkhankham Neua , Sisavat Tay and Hatsadi Neua ) . Numerous individual ( age , education level , occupation , length or residence in Vientiane , tactile contact with rats ) , household ( income , size and sanitary condition ) and neighbourhood characteristics ( level of urbanization and density of buildings ) were statistically associated in bivariate analysis with IgG antibodies against scrub typhus ( Table 3 ) . The older the population , the higher was the prevalence of scrub typhus seropositivity ( Figure 3 ) . In multivariate analysis , eleven characteristics remained significantly associated with IgG antibodies against scrub typhus . Being female , >55 years old , a farmer , Lao citizen , having had no education , had tactile contact with rats , living in households with low income , in larger households ( >4 people ) , in plots of land with poor sanitary conditions , living in neighborhoods with a low building density and having lived in Vientiane City for less than one third of lifetime increased significantly the risk for past infection with scrub typhus ( Table 4 ) . Factors associated in bivariate analysis with the presence of IgG antibodies against murine typhus were less frequent: longer residence in Vientiane , absence of recent labour in rice fields , living in central urban zone and in neighbourhoods with higher building density ( Table 3 ) . In multivariate analysis ( Table 5 ) , four characteristics were significantly associated with IgG antibodies against murine typhus - living in Vientiane for more than one third of their lifetime , being a retail trader , staying at home or not working , living close to markets ( <750 m ) and in neighborhoods with high building density increased significantly the risk for past infection with murine typhus . In multivariate analysis ( table not presented ) people aged >55 years ( OR = 2 . 5; 95%CI = 1 . 3–4 . 8; p<0 . 01 ) , those from a poor household ( OR = 2 . 3; 95%CI = 1 . 0–5 . 3; p = 0 . 05 ) and those living in plots of land with poor sanitary conditions ( OR = 1 . 8; 95%CI = 1 . 0–3 . 5; p = 0 . 05 ) were at greater risk of having IgG antibodies against both scrub and murine typhus . None of the neighborhood factors were associated with the concurrent presence of IgG antibodies against both scrub and murine typhus .
This study examined the presence of IgG antibodies as surrogate markers for past infection with agents known to cause two common rickettsial diseases in and around an Asian city . Patients with both diseases present at Vientiane health facilities and are sympatric at the district level [20] . However , adults living in central , more urbanised area of Vientiane had a higher seropositivity against murine typhus and , adults living in peripheral less urbanised Vientiane had a higher seropositivity against scrub typhus . This confirms that those living in more rural areas are at higher risk of scrub typhus infection and , those living in urban areas are at more risk of murine typhus infection , which is consistent with what has been observed elsewhere [11] , [12] , [15] , [16] , [17] , [18] , [19] . The absence of previous serological surveys does not allow examination of temporal changes in transmission of these diseases , which were first described in Laos in 2006 [20] . In a paper resulting from the same survey , variation according to the level of urbanization was also noticed for the spatial distribution of flavivirus exposure in Vientiane city with anti-flavirus IgG prevalence significantly higher among individuals living in the central city ( 60 . 1% ) than those living in the periphery ( 44 . 3% ) [29] . There are at least three possible explanations for apparent urban scrub typhus in Vientiane . Urban inhabitants may have acquired the infection in prior rural residence elsewhere , in visits to rural areas to help with farming , collecting bamboo shoots , hunting and fishing or they could have contracted the infection in urban areas . All are likely to be important . Anecdotally , a significant minority of patients admitted with scrub typhus to Mahosot Hospital , Vientiane , had been to rural areas during the putative incubation period ( ∼7–10 days ) to help with their families' farming ( RP , PN ) but the disease could also be contracted in parks , fields and gardens within the city . In part because of the terrible toll O . tsutsugamushi took on troops in scrublands in Burma & NE India in the Second World War it came to be known as scrub typhus . However , contrary to what textbooks still claim [30] scrub typhus also commonly occurs in palm plantations , primary forest , beaches , gardens [14] , [31] and also from metropolitan areas as Bombay ( Mumbai ) [32] , Jakarta [16] , [33] , [34] , suburban Bangkok [35] , [36] , Komatsu City , Japan [37] , Yuxi City , China [19] and Calcutta [38] . In view of the broader ecological distribution than is implied by the term ‘scrub typhus’ , the original Japanese name of tsutsugamushi , as suggested by Cadigan et al . [14] , may be less confusing . Richards et al . [17] examined the seroepidemiology of sympatric murine and scrub typhus in Java through a cross-sectional community-based survey in rural , suburban and urban areas in Malang District . They found prevalences of anti-O . tsutsugamushi IgG and anti-R . typhi IgG of 1 . 3% and 34 . 7% , respectively . Amongst , presumably urban , Kuala Lumpur blood donors , Tay et al . [18] found that 5 . 4% and 9 . 2% were seropositive against O . tsutsugamushi and R . typhi , respectively . They suggested that ‘with rapid economic development…the close proximity of agricultural habitats and urban development may allow the transmission of rickettsial diseases in the urban areas’ [18] . The results from Vientiane contrast with those from Malang and Kuala Lumpur in that many more inhabitants of Vientiane had evidence for prior exposure to scrub typhus . Without comparable data on rural exposure and length of residence in urban areas it is difficult to interpret this difference . Indices of poverty , such as level of education , farming , low household income and poor plot sanitary conditions were studied in relation with past exposure to scrub and murine typhus . Although these indices are not independent , it is interesting to note that they are all significantly associated with past exposure to scrub typhus in multivariate analysis . However , it is not the case for past exposure to murine typhus since , of variables potentially linked to poverty , only occupation appears as a significant variable . Different aspects of poverty may then be involved in risks to exposure to scrub or murine typhus . More qualitative research is needed to evaluate relationship between poverty and typhus transmission . That non-Lao people are at less risk to previous exposure to scrub typhus may reflect the fact that they are less involved in familial farming activities and that they are preferentially living in more urbanised neighbourhoods [23] . Why women should be at apparent increased risk to have been exposed to scrub typhus is unclear as both sexes participate in farming . The probability of having positive IgG antibodies against scrub typhus clearly increased with age . However , we did not observe any significant relationship between age and seropositivity for murine typhus . Adults with long residency in Vientiane had a higher frequency of IgG antibodies against murine typhus . As rodent and flea densities are likely to be higher in urban settings , people living for a long time in Vientiane may be more exposed to murine typhus . This is supported by the association of IgG against murine typhus with higher building density and closeness to markets and is reflected in the old name of ‘shop typhus’ for murine typhus . Limitations of this study include that we did not collect information about other places visited outside residential neighbourhood ( i . e . place of work , leisure , travel ) or former places of residence , and that we used old land use coverage data whilst the Vientiane landscape has changed considerably since these land use data were collected in 1999 . In addition only one person was sampled in each household , which did not allowed for examination of clustering within households . Furthermore we surveyed only adults >35 years old and could therefore not examine changes in seroprevalence in younger people , in whom a significant proportion of seroconversions are likely to occur [16] , [17] . It is unknown what proportion of patients with IgG against these two rickettsial agents developed disease . Surprisingly , there are few data on the longevity of IgG , on locally appropriate cut offs indicating past infection or of the consequences of repeated infections on antibody titres against O . tsutsugamushi and none that we are aware of for IgG against R . typhi . Saunders et al . [39] calculated that the annual reversion rate for antibodies against O . tsutsugamushi to a titre <l:50 was 6l% , suggesting that our data are likely to underestimate the true frequency of past infections . Hence , there remains uncertainty as to the most appropriate diagnostic techniques and correct serology cutoff titres in different environments in the diagnosis of prior rickettsial infections in the healthy . The presence of IgG against these pathogens should be regarded as indices of exposure . Some patients may develop specific IgG without developing disease and some will lose their specific IgG during their lifetime . More research is needed to define antibody responses , against both pathogens , and their changes through time to define both acute disease [40] and past exposure . Possible cross-reactivity of the murine typhus IgG ELISA with other rickettsial pathogens is not entirely excluded but the most likely candidate for this , R . prowazekii , has not been recorded from Laos . Vientiane is a rapidly growing city , with the population having almost doubled during the past 15–20 years [25] . As a result , the territory of the city has expanded into the paddy fields of its former rural hinterland and the city has embarked on a far-reaching path of urban transformation . The transformation in physical landscape and in way of life may have and will lead to a modification of scrub and murine transmission . Both scrub and murine typhus tend to afflict the poorer citizens of Vientiane and are responsible for high incidence of undifferentiated fever . As they are relatively simple and inexpensive to treat , oral doxycycline may be an appropriate empirical therapy for those for without access to confirmatory tests . However , it would need to be borne in mind that other common diseases , from which rickettsial diseases are difficult to distinguish clinically , such as typhoid and dengue , would not respond to such therapy . Public education campaigns on disease avoidance and chigger repellents , and community participation ( rubbish disposal and rodent control , especially at markets ) may reduce the incidence of scrub typhus and murine typhus [41] , [42] . More knowledge is needed on the vectors and epidemiology of rickettsial diseases in urban Laos , especially whether scrub typhus-infected chiggers occur in urban areas and which flea and rodent species are involved in the local epidemiology of murine typhus .
|
Scrub typhus and murine typhus are neglected but important treatable causes of fever , morbidity and mortality in South-East Asia . Epidemiological data suggests that scrub typhus would be more common in rural areas and murine typhus in urban areas but there are very few comparative data from places where both diseases occur , as is the case in Vientiane , the capital of the Lao PDR . We therefore determined the frequency of IgG antibody seropositivity against scrub typhus and murine typhus , as indices of prior exposure to these pathogens , in a randomly selected population of 2 , 002 adults living in different neighbourhoods in Vientiane . The overall prevalence of IgG against these two pathogens was ∼20% . However , within the city , the spatial distribution of IgG against these two diseases was radically different - past exposure to murine typhus being more frequent in urbanized areas while past exposure to scrub typhus more frequent in outlying areas . This study underscores the importance of ecological characteristics in improving the understanding of both scrub typhus and murine typhus transmission and epidemiology .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"public",
"health",
"and",
"epidemiology/epidemiology",
"public",
"health",
"and",
"epidemiology/social",
"and",
"behavioral",
"determinants",
"of",
"health",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"public",
"health",
"and",
"epidemiology/environmental",
"health"
] |
2010
|
Contrasting Spatial Distribution and Risk Factors for Past Infection with Scrub Typhus and Murine Typhus in Vientiane City, Lao PDR
|
Asymmetrical dimethylarginine ( ADMA ) , an endogenous inhibitor of nitric oxide synthase ( NOS ) , is a predictor of mortality in critical illness . Severe malaria ( SM ) is associated with decreased NO bioavailability , but the contribution of ADMA to the pathogenesis of impaired NO bioavailability and adverse outcomes in malaria is unknown . In adults with and without falciparum malaria , we tested the hypotheses that plasma ADMA would be: 1 ) increased in proportion to disease severity , 2 ) associated with impaired vascular and pulmonary NO bioavailability and 3 ) independently associated with increased mortality . We assessed plasma dimethylarginines , exhaled NO concentrations and endothelial function in 49 patients with SM , 78 with moderately severe malaria ( MSM ) and 19 healthy controls ( HC ) . Repeat ADMA and endothelial function measurements were performed in patients with SM . Multivariable regression was used to assess the effect of ADMA on mortality and NO bioavailability . Plasma ADMA was increased in SM patients ( 0 . 85 µM; 95% CI 0 . 74–0 . 96 ) compared to those with MSM ( 0 . 54 µM; 95%CI 0 . 5–0 . 56 ) and HCs ( 0 . 64 µM; 95%CI 0 . 58–0 . 70; p<0 . 001 ) . ADMA was an independent predictor of mortality in SM patients with each micromolar elevation increasing the odds of death 18 fold ( 95% CI 2 . 0–181; p = 0 . 01 ) . ADMA was independently associated with decreased exhaled NO ( rs = −0 . 31 ) and endothelial function ( rs = −0 . 32 ) in all malaria patients , and with reduced exhaled NO ( rs = −0 . 72 ) in those with SM . ADMA is increased in SM and associated with decreased vascular and pulmonary NO bioavailability . Inhibition of NOS by ADMA may contribute to increased mortality in severe malaria .
Plasmodium falciparum causes ∼1 million deaths annually [1] , [2] . Despite rapid parasite clearance with the anti-parasitic drug artesunate , the mortality rate in severe malaria remains high [3] , [4] . Endothelial activation , parasite sequestration , impaired microvascular perfusion and dysregulated inflammatory responses are all thought to contribute to severe and fatal malaria [5]–[9] . Increased understanding of these pathogenic mechanisms may identify targets for adjunctive therapies to further improve outcomes . Severe malaria is associated with impaired nitric oxide ( NO ) bioavailability and blood mononuclear cell NO synthase ( NOS ) type 2 expression in both children [10] , [11] and adults [6] . The concentrations of L-arginine , the substrate for NO production by all three NOS isoforms [12] , are low in children and adults with severe malaria and likely contribute to the decreased NO production found in severe disease [6] , [10] , [13] . However , in adults with moderately severe malaria , L-arginine concentrations are at least as low as those seen with severe malaria , yet there is no impairment of vascular and pulmonary NO bioavailability as found in severe disease [6] . This suggests that factors other than substrate limitation contribute to impaired NO bioavailability in severe malaria . Asymmetrical dimethylarginine ( ADMA ) is a non-specific endogenous NOS inhibitor which decreases vascular function in cardiovascular and renal disease [14] , [15] . Protein-arginine-methyltransferases methylate arginine residues in proteins and ADMA is released when these proteins undergo degradation [14] . ADMA is primarily eliminated by the enzyme dimethylarginine-dimethylaminohydrolase-1 ( DDAH-1 ) in the liver and kidney , with ∼20% being excreted in the urine [16] . In adult sepsis , elevated ADMA is independently associated with increased mortality , a likely consequence of non-specific inhibition of homeostatic NO production [17] , [18] . Increased protein catabolism with hepatic and renal dysfunction in severe malaria has the potential to increase ADMA and impair NO production , but the importance of ADMA in the pathogenesis of malaria is currently unknown . Clarification of the role of ADMA in malaria is of particular importance given a recent genome-wide association study in children linking DDAH polymorphisms with risk of severe malaria [19] , and the potential for the parasite as well as host to produce ADMA [20] . Acute lung injury is a common but little-studied complication of severe falciparum malaria associated with high mortality [21] . In sepsis and critical illness , acute lung injury and mortality are associated with decreased total and pulmonary NO [22] , [23] . Pulmonary diffusion capacity and exhaled NO concentrations are both reduced in severe malaria [6] , [24] , however the causative factors have not been identified . The role of ADMA in impairing pulmonary NO bioavailability in severe malaria , or indeed any critical illness , is not known . In a prospective longitudinal study of Indonesian adults with malaria , we evaluated the hypotheses that concentrations of methylated arginines are independently associated with a ) disease severity , b ) reduced exhaled NO and vascular NO bioavailability and c ) increased mortality .
We measured asymmetrical dimethylarginine ( ADMA ) , symmetrical dimethylarginine ( SDMA ) and L-arginine in 49 patients with severe malaria , 78 with moderately severe malaria and 19 healthy controls . In the SM patients , 20 patients had only one criterion for severe disease ( coma in 14 , hyperbilirubinemia in 4 , respiratory distress in 2 ) with the remainder having >1 criteria . In total , 34 of the patients with SM were treated with intravenous artesunate and the remaining 15 received intravenous quinine [6] . All of the 78 MSM patients were treated with quinine with the exception of one who received artesunate . Exhaled NO concentrations ( FeNO ) could not be measured in those with coma , but were possible in 48% ( 11/23 ) of non-comatose SM patients , 88% ( 69/78 ) of MSM patients and 100% ( 19/19 ) of HCs . RH-PAT index was measured in all patients with malaria as well as HC . There were eight deaths among the patients with SM , and none in the MSM patients . Repeat RH-PAT and venous blood measurements were only performed in one and four of the eight fatal cases respectively . Baseline characteristics of study participants are summarized in Table 1 . ADMA and SMDA concentrations were increased in SM patients ( 0 . 85 µM; 95% CI 0 . 74–0 . 96 and 1 . 67 µM; 95% CI 1 . 24–2 . 09 respectively ) compared to those with MSM ( 0 . 54 µM; 95%CI 0 . 5–0 . 56 and 0 . 58 µM; 95%CI 0 . 54–0 . 63 ) and HC ( 0 . 64 µM; 95%CI 0 . 58–0 . 70 and 0 . 53 µM; 95%CI 0 . 47–0 . 59 ) ; ANOVA p<0 . 001 for both ADMA and SDMA , Table 2 , Figure 1A and 1B . L-arginine concentrations were significantly higher in HCs compared to patients with MSM and SM , with the L-arginine/ADMA ratio decreasing with increasing disease severity ( p<0 . 001; Table 2 ) . Exhaled nitric oxide concentration ( FeNO ) was significantly decreased in SM patients compared to MSM and HC; ( Table 2 ) . FeNO was inversely correlated with plasma ADMA concentration ( rs = −0 . 31 , p = 0 . 003; Table 3 ) in all malaria patients , and in the subgroup of 11 patients with SM ( rs = −0 . 72 , p = 0 . 01; Table 3 ) . FeNO also correlated inversely with HRP2 concentration ( rs = −0 . 51 , p<0 . 001 ) in malaria patients , but not in the SM group . FeNO remained inversely associated with ADMA , after adjusting for disease severity , creatinine and HRP2 . FeNO was not associated with SDMA , the L-arginine/ADMA ratio , plasma hemoglobin or arginase . As reported previously , the RH-PAT index was significantly lower in SM patients compared to those with MSM and controls ( p<0 . 001 ) Table 2 . In all malaria patients , there were moderate inverse associations between RH-PAT index and ADMA ( rs = −0 . 32; p<0 . 001; Table 3 ) and SDMA ( rs = −0 . 35; p<0 . 001 ) concentrations . After adjusting for factors previously shown to be associated with RH-PAT index including plasma hemoglobin [6] , [7] , [25] , the inverse association with ADMA remained significant , with the final model including both ADMA and cell free hemoglobin ( r = 0 . 40 ) . In contrast , the L-arginine/ADMA ratio was not associated with the RH-PAT index . Longitudinally there was no association between the RH-PAT index and ADMA concentration or L-arginine/ADMA ratio in SM patients . Patients with SM had significantly elevated plasma concentrations of Ang-2 , ICAM-1 and E-selectin compared to those with MSM and HC; Table 2 . Angiopoietin-2 and ICAM-1 were significantly correlated with both ADMA ( rs = 0 . 48 and 0 . 42 respectively; p<0 . 001; Table 3 ) , and SDMA ( rs = 0 . 54 and 0 . 52; p<0 . 001 ) , and this was also apparent in the subgroup of SM patients; Table 3 . ADMA remained independently associated with Ang-2 and ICAM-1 after adjusting for confounding factors , including creatinine , plasma hemoglobin , parasite biomass and disease severity . There was no significant correlation between ADMA or SDMA with E-selectin , and none between the L-arginine/ADMA ratio and Ang-2 , ICAM-1 and E-selectin . The plasma creatinine , total bilirubin , P . falciparum histidine rich protein 2 ( HRP2 ) and venous lactate were increased in SM compared to MSM ( Table 2 ) . In all patients with malaria , there were correlations between ADMA and SDMA with creatinine ( rs = 0 . 45; p<0 . 001; rs = 0 . 69; p<0 . 001; Table 3 ) , total bilirubin ( rs = 0 . 32; p<0 . 001; rs = 0 . 36; p<0 . 001; Table 3 ) , HRP2 ( rs = 0 . 46; p<0 . 001; rs = 0 . 62; p<0 . 001; Table 3 ) and lactate ( rs = 0 . 3; p = 0 . 01; rs = 0 . 29 , p = 0 . 02; Table 3 ) . The associations remained significant in SM patients for each of these biomarkers of disease severity except for venous lactate ( Table 3 ) . There was no association between L-arginine/ADMA ratio and any biomarker . TNF was only measured in the SM patients , and was associated with SDMA ( rs = 0 . 52; p = 0 . 001 ) , but not with ADMA ( p = 0 . 06 ) . In SM patients , ADMA and SDMA concentrations were significantly higher in the 8 patients who died ( 1 . 28 µM; 95%CI 0 . 88–1 . 74 and 3 . 76 µM; 95%CI 1 . 88–5 . 56 , respectively ) compared to the 41 survivors ( 0 . 77 µM; 95%CI 0 . 64–0 . 84 and 1 . 27 µM; 95%CI 0 . 99–1 . 54 , respectively; p<0 . 001; Figure 1A and 1B ) . Each micromolar increase in ADMA and SDMA concentrations was associated with an 18-fold ( OR 18 . 8 95% CI 2 . 0–181; p = 0 . 01 ) and three-fold ( OR 3 . 0; 95% CI 1 . 5–6 . 2; p = 0 . 002 ) increased risk of death , respectively . ADMA but not SDMA remained a significant risk factor for death after adjusting for other confounding factors , such as Ang-2 , creatinine , parasite biomass , bilirubin , base deficit and lactate . A final model predicting a fatal outcome included ADMA , Ang-2 , HRP2 and creatinine ( Table 4 ) . The L-arginine/ADMA ratio was not associated with risk of death . The prognostic value of ADMA in predicting a fatal outcome was measured by the area under the receiver operating curve ( ROC ) . ADMA ( AUROC 0 . 85; 95% CI 0 . 71–0 . 99; Figure 2 ) was comparable to other reliable prognostic indicators , including Ang-2 ( AUROC 0 . 84; 95% CI 0 . 71–0 . 96 ) , HRP2 ( AUROC 0 . 86; 95% CI 0 . 73–0 . 94 ) , base deficit ( AUROC 0 . 73; 95% CI 0 . 53–0 . 92 ) , and TNF ( AUROC 0 . 71; 95% CI 0 . 43–0 . 98 ) , and a better predictor of fatal outcome than venous blood lactate ( AUROC 0 . 63; 95%CI 0 . 41–0 . 83; p = 0 . 003 ) . In patients with severe malaria , there was no significant change in ADMA ( Figure 3A ) or SDMA concentrations during the course of hospitalization among the overall group , survivors or those with a fatal outcome . Among survivors , there was a daily increase in L-arginine/ADMA ( β = 9 . 1 , p<0 . 001; Figure 3B ) but no increase in those who died .
ADMA is increased in severe falciparum malaria and is an independent predictor of mortality . Indeed ADMA was a better predictor of death than blood lactate , previously shown to be a reliable prognostic indicator of increased mortality in severe malaria [26] . Our study demonstrated that elevated plasma ADMA concentrations are independently associated with decreased exhaled NO concentrations , impaired vascular NO bioavailability , increased endothelial activation and parasite biomass . To our knowledge this is the first demonstration of a relationship between increased ADMA and impaired exhaled NO in any critical illness . Taken together , these findings suggest that ADMA , an endogenous inhibitor of all three nitric oxide synthase ( NOS ) isoforms , reduces NO bioavailability in at least two organ systems and may contribute to increased mortality in falciparum malaria . In critically ill patients , elevated ADMA concentrations are likely to result from increased production and reduced elimination . The elevation in both ADMA and SDMA may result from increased host production of methylated arginines in severe malaria . The majority of circulating ADMA is taken up by the liver before being metabolized by dimethylarginine-dimethylaminohydrolase-1 ( DDAH-1 ) ; approximately 20% is excreted unchanged in the urine [16] . Hepatic blood flow is known to be significantly impaired in severe malaria [27] . The correlation of bilirubin and creatinine with increased ADMA , suggests that similar to sepsis [28] , decreased hepatic and renal elimination may also increase ADMA concentrations in severe malaria . The large parasite biomass in severe malaria may also be a potential source of ADMA , with Plasmodium falciparum possessing protein arginine methyltransferases capable of producing ADMA [20] . The significant independent correlation between parasite biomass and ADMA on admission suggests this may be occurring in vivo , although the persistently elevated levels in severe malaria after commencement of anti-malarial therapy suggest the importance of altered host production and clearance in the post-treatment period . Increased clearance due to increases in either hepatic blood flow or DDAH activity may explain the decreased ADMA concentrations in patients with moderately severe malaria . There are no clinical studies to date documenting increased DDAH activity in mild inflammatory diseases , but hepatic blood flow is known to be significantly increased in acute uncomplicated falciparum malaria compared to patients with severe disease and healthy individuals [27] . The converse may also be true , with lower plasma ADMA concentrations potentially increasing vascular NO bioavailability in moderately severe malaria [14] , and possibly contributing to elevated hepatic blood flow . The loss of DDAH-1 function in a murine model of sepsis increased ADMA , reduced NO signaling , and worsened vascular pathophysiology including endothelial function [29] . In human severe sepsis , a polymorphism in the DDAH-2 enzyme increased ADMA levels which were associated with increased severity of organ failure and early septic shock [30] . Recently , a genome-wide association study in children found that a polymorphism in the gene encoding DDAH-1 was associated with an increased likelihood of severe malaria [19] . These studies indicate that altered DDAH function may be a contributor to organ damage and increased mortality in severe malaria as well as in other critical illnesses . SDMA does not inhibit NOS , but competes with plasma L-arginine for intracellular uptake by the cationic amino acid transporters ( CAT ) . Unlike ADMA , it is not metabolized by DDAH and is almost exclusively eliminated by the kidneys [16] . In chronic disease SDMA has recently been shown to be an independent predictor for major cardiovascular events in certain chronic diseases [31] . We find that in malaria , SDMA concentrations are associated with mortality and decreased vascular bioavailability on univariate analysis , but not after adjusting for renal function . While the association between SDMA and disease severity is likely to reflect the degree of renal impairment and SDMA retention , it is possible that retained SDMA may also contribute to decreased NO bioavailability in severe malaria . In critically ill adults with organ failure and severe sepsis , ADMA concentrations are associated with increased all-cause mortality and the severity of organ failure [17] , [30] . Investigators have hypothesized that this may result from non-selective inhibition by ADMA of all three isoforms of NOS , particularly homeostatic NOS3 ( endothelial NOS ) [18] . This is similar to the postulated mechanism to explain the increased mortality with use of NG-monomethyl-arginine ( NMMA ) , another non-specific NOS inhibitor , in a phase 3 clinical trial of sepsis [32] . In falciparum malaria , systemic NO production is impaired in severe disease and hypoargininemia is likely to be a contributing cause [6] , [10] , [11] , [13] . Exhaled and vascular NO are both reduced in adults with severe malaria [6] , but not in moderately severe malaria ( MSM ) despite similar degrees of hypoargininemia [6] . This may be explained by the higher ADMA in severe malaria and a greater competitive inhibition of NOS in SM compared to MSM , similar to clinical studies of healthy volunteers in which ADMA infusion reduced blood flow [15] , [33] . In mouse studies , ADMA infusion alone reduces splenic blood perfusion , but when combined with hypoargininemia , causes a reduction in renal , hepatic and splenic blood flow with organ damage [34] . Regulation of microcirculatory flow is dependent on pre-capillary arteriolar vasodilatory responses which in turn are critically dependent on NO production [35] , with both likely to be decreased by ADMA in SM . By decreasing functional capillary density , ADMA could further impair microcirculatory function already compromised by parasite sequestration in capillaries and post-capillary venules [36] . We have previously shown that hemolysis-related NO quenching by cell-free hemoglobin is associated with reduced vascular NO bioavailability in severe malaria [25] . In malaria , increased ADMA and cell-free hemoglobin were independently related to endothelial dysfunction , suggesting that inhibition of NOS and NO quenching both reduce vascular NO bioavailability . NO has multiple regulatory functions that maintain endothelial quiescence in vitro , including inhibition of endothelial Weibel-Palade body ( WPB ) exocytosis and ICAM-1 expression [37] , [38] . Plasma concentrations of angiopoietin-2 ( Ang-2 ) , an angiogenic factor stored in WPBs , predict increased mortality in malaria [7] , and ICAM-1 is a major endothelial adhesion receptor mediating cytoadherence of parasitized red cells and microvascular sequestration [5] . We demonstrate that ADMA levels correlate with increased Ang-2 , but the association between ADMA and increased mortality is independent of Ang-2 , suggesting effects of NO inhibition in addition to increased WPB exocytosis . Acute lung injury is a complication of severe malaria in adults associated with a high mortality rate [21] , [24] . Gas transfer at the alveolar-capillary membrane and exhaled NO are both decreased in severe falciparum malaria [6] , [24] . In an animal model of sepsis–associated pulmonary injury , non-selective NOS inhibition causes increased lung edema [39] . Clinical studies have shown decreased pulmonary NO concentrations in patients with acute respiratory distress syndrome , as well as an association between decreased NO production and a worse outcome in acute lung injury [22] , [23] . The lung is a major source of ADMA and increased concentrations are associated with pulmonary arterial hypertension [40] , [41] . In severe malaria , both ADMA and parasite biomass are strongly inversely associated with exhaled NO concentrations , suggesting that both factors impair pulmonary NO production in severe disease . There are several limitations in our study . Measurement of exhaled NO was not possible in patients with coma and was possible in only half of severe malaria patients without coma . Our results may therefore not reflect the relationship between ADMA and exhaled NO in all syndromes of severe malaria . In patients who died , only 4 of 8 had at least one repeat blood sample , and the longitudinal data may not truly reflect the course of the methylated arginines in fatal cases . RH-PAT index is at least 50% dependent on endothelial NO release [42] , but we cannot exclude an effect of ADMA on other vasodilators such as prostacyclin and endothelium-derived hyperpolarizing factor . Although we have measured plasma ADMA concentrations , the effects of ADMA are intracellular . Nevertheless , in vitro studies with endothelial cells have shown that increasing extracellular ADMA results in five-fold increases in intracellular concentrations . This suggest that intracellular concentrations of ADMA in severe malaria may be higher , and may be adequate for meaningful inhibition of for all three NOS isoforms ( IC50s ∼2-5µM ) [43] . The observational nature of the study does not allow us to conclude with certainty a direct role for ADMA in the pathophysiology of severe malaria . While the association of ADMA with mortality may reflect impaired renal and hepatic function , it remained significant after adjusting for these factors in a multivariable model . Furthermore , increased ADMA from impaired hepatic and/or renal clearance is not just a marker of organ dysfunction in critical illness , with retained ADMA having functional consequences on NOS activity . In summary , the endogenous non-selective NOS inhibitor ADMA is elevated in SM and is an independent risk factor for mortality . ADMA is also associated with decreased FeNO and vascular NO bioavailability , as well as increased endothelial activation and parasite biomass . Therapies which increase NO bioavailability or which diminish ADMA levels represent rational approaches for interventional trials of adjunctive therapy in severe malaria .
The study was conducted at Mitra Masyarakat Hospital , Timika , Papua , Indonesia , a region with unstable transmission of multidrug resistant malaria [44] , [45] . Written informed consent was obtained from all patients , if they were comatose or too ill , consent was obtained from relatives . The Ethics Committees of the National Institute of Health Research and Development , Indonesia , and Menzies School of Health Research , Australia approved the study . Patients were ≥18 years old with moderately-severe ( MSM ) or severe ( SM ) Plasmodium falciparum malaria without P . vivax infection and with a hemoglobin level >60 g/L who had been prospectively enrolled in a study of endothelial dysfunction and exhaled NO [6] . Previous results from this study group have been published [6] , [7] , [13] , [25] . Briefly , SM was defined as P . falciparum parasitemia and ≥1 modified WHO criterion of severity ( excluding severe anemia ) . MSM was defined as fever within the preceding 48 hours , >1 , 000 asexual P . falciparum parasites/µL , no WHO warning signs or severe malaria criteria and a requirement for inpatient parenteral therapy because of inability to tolerate oral treatment . Healthy controls ( HC ) were non-related hospital visitors with no history of fever in last 48 hours , intercurrent illness or smoking in last 12 hours , or evidence of parasitemia [6] . Standardized history and physical examination were documented . Heparinized blood was collected daily , centrifuged within 30 minutes of collection and plasma stored at −70°C . Parasite counts were determined by microscopy . Hemoglobin , biochemistry , acid-base parameters and lactate were measured with a bedside analyser ( i-STAT Corp ) . Patients were treated with anti-malarials and antibiotics using standard national protocols as previously described [6] . Solid phase extraction ( SPE ) of amino acids was followed by derivatisation with Accq-Fluor and separation on a Gemini-NX column at pH 9 [46] . The SPE method gives absolute recoveries of >80% for ADMA and symmetrical dimethylarginine ( SDMA ) and average relative recoveries of 102% for ADMA and 101% for SDMA . The HPLC method gives intra-assay RSDs of 2 . 1% and 2 . 3% and inter-assay RSDs of 2 . 7% and 3 . 1% for ADMA and SDMA respectively [46] . Plasma concentrations of cell-free hemoglobin and the endothelial activation markers , ICAM-1 , E-selectin and angiopoietin-2 were measured by ELISA as previously reported in this population [6] , [7] , [25] . Total parasite biomass was quantified by measuring plasma histidine rich protein-2 ( HRP2 ) by ELISA [6] , plasma arginase by a radiometric method [6] and plasma TNF concentrations by flow cytometry , as previously reported [7] . Endothelial function was measured non-invasively using peripheral arterial tonometry ( EndoPAT ) by the change in digital pulse wave amplitude in response to reactive hyperemia , giving a RH-PAT Index as reported previously [6] . The RH-PAT index is at least 50% dependent on endothelial NO production [42] . Endothelial function was measured daily until death or discharge , or until the RH-PAT index was above an a priori cutoff ( 1 . 67 ) for two consecutive days [13] . Fractional concentrations of exhaled NO were measured by NO analyser ( Aerocrine ) , as previously described , using American Thoracic Society guidelines and a flow rate of 250 ml/sec [6] . Statistical analysis was performed with STATA 9 . 2 software . Intergroup differences were compared by ANOVA or Kruskal-Wallis test , where appropriate . Pearson's or Spearman's correlation coefficients were determined depending on normality of distributions . Multiple stepwise linear regression was used to adjust for confounding variables . Longitudinal associations were assessed by mixed effects modeling using generalized estimating equations . Logistic regression was used to determine the association between death and ADMA concentrations . Variables hypothesized , as well as those shown in previous publications [6] , [7] , [25] , to contribute to mortality , pulmonary NO and endothelial pathology were included in a multiple regression model if p<0 . 05 on univariate analysis and retained if they remained significant . Goodness-of-fit was assessed by the Hosmer-Lemeshow goodness of fit test and independent variables tested for interactions . The prognostic utility of continuous variables was measured using the area under the receiver operating curves ( ROCs ) and its 95% confidence intervals were calculated . A two-sided value of p<0 . 05 was considered significant .
|
Severe falciparum malaria is associated with impaired microvascular perfusion , lung injury and decreased bioavailability of nitric oxide ( NO ) , but the causes of these processes are not fully understood . Asymmetrical dimethylarginine ( ADMA ) , a competitive endogenous inhibitor of nitric oxide synthase ( NOS ) , is an independent predictor of mortality in other critical illnesses , and can impair vascular function in chronic disease . ADMA can be produced by both the host and malaria parasites . The major novel findings of this study in malaria are that ADMA is an independent predictor of death in falciparum malaria , and is associated with decreased availability of nitric oxide in at least two organ systems affected by malaria parasites , the lining of blood vessels and the lungs . This study contributes to knowledge of regulation and availability of pulmonary and endothelial NO in critical illness and identifies pathogenic processes which may contribute to death in severe malaria . Therapies which increase the availability of NO or which reduce ADMA levels may have potential for adjunctive therapy of severe malaria .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"immunology/cellular",
"microbiology",
"and",
"pathogenesis",
"infectious",
"diseases/protozoal",
"infections"
] |
2010
|
Increased Asymmetric Dimethylarginine in Severe Falciparum Malaria: Association with Impaired Nitric Oxide Bioavailability and Fatal Outcome
|
Coxiella burnetii is an intracellular bacterium that replicates within an expansive phagolysosome-like vacuole . Fusion between the Coxiella-containing vacuole ( CCV ) and late endosomes/multivesicular bodies requires Rab7 , the HOPS tethering complex , and SNARE proteins , with actin also speculated to play a role . Here , we investigated the importance of actin in CCV fusion . Filamentous actin patches formed around the CCV membrane that were preferred sites of vesicular fusion . Accordingly , the mediators of endolysosomal fusion Rab7 , VAMP7 , and syntaxin 8 were concentrated in CCV actin patches . Generation of actin patches required C . burnetii type 4B secretion and host retromer function . Patches decorated with VPS29 and VPS35 , components of the retromer , FAM21 and WASH , members of the WASH complex that engage the retromer , and Arp3 , a component of the Arp2/3 complex that generates branched actin filaments . Depletion by siRNA of VPS35 or VPS29 reduced CCV actin patches and caused Rab7 to uniformly distribute in the CCV membrane . C . burnetii grew normally in VPS35 or VPS29 depleted cells , as well as WASH-knockout mouse embryo fibroblasts , where CCVs are devoid of actin patches . Endosome recycling to the plasma membrane and trans-Golgi of glucose transporter 1 ( GLUT1 ) and cationic-independent mannose-6-phosphate receptor ( CI-M6PR ) , respectively , was normal in infected cells . However , siRNA knockdown of retromer resulted in aberrant trafficking of GLUT1 , but not CI-M6PR , suggesting canonical retrograde trafficking is unaffected by retromer disruption . Treatment with the specific Arp2/3 inhibitor CK-666 strongly inhibited CCV formation , an effect associated with altered endosomal trafficking of transferrin receptor . Collectively , our results show that CCV actin patches generated by retromer , WASH , and Arp2/3 are dispensable for CCV biogenesis and stability . However , Arp2/3-mediated production of actin filaments required for cargo transport within the endosomal system is required for CCV generation . These findings delineate which of the many actin related events that shape the endosomal compartment are important for CCV formation .
Coxiella burnetii is an intracellular Gram-negative bacterium that causes a severe flu-like disease called Q fever . Q fever is a zoonosis , and transmission to humans typically occurs by inhalation of contaminated aerosols generated by infected ruminant livestock [1] . Following deposition in the lungs , C . burnetii invades and replicates within alveolar macrophages [2] . C . burnetii grows within an expansive phagolysosome-like compartment , termed the Coxiella-containing vacuole ( CCV ) , that is acidic and displays robust hydrolytic activity [3 , 4] . Enlargement of the nascent phagosome to accommodate pathogen growth primarily occurs through fusion with late endosomes/multivesicular bodies , although interactions also occur with autophagic and secretory pathways [5–11] . Endocytic maturation of the CCV and fusion with lysosomes requires sequential engagement of the GTPases Rab5 and Rab7 , with Rab7 binding the HOPS ( homotypic fusion and vacuole protein sorting ) tethering complex [5 , 12 , 13] . This interaction promotes vesicular fusion by trans-SNARE ( soluble N-ethymaleimide-sensitive factor-attachment protein receptor ) complexes , consisting of vesicle-associated membrane protein 7 ( VAMP7 ) , Vti1b , syntaxin 7 , and syntaxin 8 [10 , 14] . Endocytic trafficking to the CCV is further facilitated by clathrin-dependent vesicular transport [15 , 16] . An siRNA screen revealed a role for the SNARE syntaxin-17 in mediating homotypic fusion of multiple CCV [11] . Secretion by C . burnetii of proteins with effector functions by a specialized Dot/Icm type 4B secretion system ( T4BSS ) is essential for CCV biogenesis , with effectors currently identified that modulate clathrin-mediated vesicular trafficking and autophagosome interactions [8 , 15–18] . Several intracellular pathogens exploit the host cytoskeleton for generation , maintenance , and/or stability of their respective replication vacuoles [19–24] . Cytoskeletal requirements for CCV formation are poorly defined , but some information exists on the role of filamentous ( F- ) actin . Colonne and co-workers [25] described an actin ring circumscribing the mature CCV in THP-1 macrophages that labels with vasodilator-stimulated phosphoprotein ( VASP ) , an actin regulatory protein . Depletion of VASP disrupts CCV formation and heterotypic fusion with latex bead-containing phagosomes . Colonne at al . [25] propose that CCV-associated F-actin stabilizes the vacuole , and that actin rearrangements enable vacuole expansion . In HeLa cells , Aguilera and co-workers [26] observed CCV association with F-actin , either as an encompassing ring or in patches . Cells treated with latrunculin B , which disrupts F-actin , exhibit aberrantly small CCVs that fully mature through the endocytic pathway but cannot fuse with latex bead-containing phagosomes . CCVs decorate with the actin-associated small GTPases RhoA and Cdc42 . Cells ectopically expressing a dominant-negative form of RhoA display a small , multi-vacuole phenotype . Aguilera et al . [26] propose that CCV F-actin participates in membrane transport events . Actin polymerization regulates late and early stages of endocytic and autophagic trafficking , respectively [27–29] . Important actin regulatory proteins directing polymerization reside in the Wiskott-Aldrich Syndrome protein ( WASP ) family [30] . These actin nucleation promoting factors ( NPFs ) activate the actin related protein-2/3 ( Arp2/3 ) complex to nucleate production of barbed actin filaments [30] . The WASP family member Wiskott-Aldrich Syndrome protein and SCAR homolog ( WASH ) also regulates endosome fission required for recycling of receptors from endosomes to the Golgi apparatus or plasma membrane . Mammalian receptor cargo is sequestered by a WASH-interacting heterotrimeric cargo-selective complex , or retromer , comprised of VPS26A or VPS26B , VPS29 , and VPS35 [29 , 31–33] . Given that the actin cytoskeleton participates in vesicular trafficking associated with CCV biogenesis , we sought to define the origin and function of CCV F-actin patches . Actin patch formation was a C . burnetii-driven process . Patches decorated with several proteins involved in fusion of late endosomes , and accordingly , these were preferred sites of CCV fusion . Actin patches also labeled with retromer and WASH complex components , as well as Arp3 . Interestingly , inhibition of retromer function reduced actin patches without affecting CCV generation and C . burnetii growth . Additionally , C . burnetii grew normally in WASH-knockout mouse embryo fibroblasts ( MEFs ) , where CCVs are devoid of actin patches . In contrast , inhibition of the Arp2/3 complex strongly inhibited CCV formation via a process involving endosomal trafficking . These data show that CCV actin patches are not required for vacuole fusion with endocytic vesicles and maintenance of the compartment . However , Arp2/3-driven actin nucleation events that regulate transport through the endosomal system are essential .
Investigating the role of actin in CCV biogenesis revealed patches of F-actin juxtaposed with the CCV membrane ( Fig 1A , S1A Fig ) . CCV actin patches have been previously described [26] , but their association with the CCV and function during infection is unclear . Formation and enlargement of the CCV occurs primarily through fusion with late endocytic vesicles [34] . To investigate the role of actin patches in CCV biogenesis , Vero and THP-1 cells were fixed at 3 days post-infection ( dpi ) , a time of accelerated vacuole growth , and stained for F-actin and markers of late endosomes ( CD63+ ) and lysosomes ( LAMP1+ ) . CCV actin patches were associated with CD63+ and LAMP1+ vesicles that frequently appeared as clusters . Large , individual vesicles were also observed with patches ( Fig 1A and 1B , S1A Fig ) . In contrast , minimal colocalization was observed between CCV actin patches and EEA1 , a marker of early endosomes ( Fig 1A and 1B ) . The interaction between CCV actin patches and late endocytic vesicles suggested that actin patches function in endosomal fusion . Therefore , localization between CCV actin patches and late endosome fusion proteins was assessed . Rab7 , a GTPase involved in regulating late endocytic membrane trafficking and fusion , localizes on the CCV membrane and is required for CCV formation and pathogen growth [11 , 35] . Staining for Rab7 in Vero cells at 3 dpi revealed Rab7 puncta on the CCV localized with F-actin patches ( Fig 1C and 1D ) . The late endocytic SNAREs , VAMP7 and its partner syntaxin 8 , also associate with the CCV membrane , with the former essential for CCV growth [36 , 37] . Similar to Rab7 , both VAMP7 and syntaxin 8 formed clusters that localized with CCV actin patches , in contrast to VAMP8 , a v-SNARE previously shown to be absent on developed CCVs [36] ( Fig 1C and 1D ) . Comparable results were seen in THP-1 cells ( S1B Fig ) . Thus , actin patches of expanding CCV accumulate proteins predicted to promote fusion between the CCV and late endosomes . To examine actin patches of more mature CCVs , Vero cells were stained for F-actin and CD63 or VAMP7 at 5 dpi , a time when C . burnetii is entering stationary phase growth and the CCV has finished expanding [38] . CCV actin patches colocalized with CD63 and VAMP7 , but compared to 3 dpi , CCVs had smaller actin patches with decreased CD63 and VAMP7 intensity ( S2A–S2E Fig ) . The role of actin in membrane fusion is not defined; however , actin polymerization is associated with tethering late endocytic vesicles to target membranes [39] . Insight into CCV actin patch function can be gained by live cell imaging . Fusion between endocytic vesicles and CCV actin patches was examined by spinning disk confocal fluorescence microscopy . CellLight reagents were used to express fluorescent proteins that label lysosomes ( LAMP1-GFP ) and actin ( actin-RFP ) . In Vero cells at 3 dpi , live cell imaging revealed docking of LAMP+ vesicles with CCV actin immediately before fusion ( Fig 2A , S1 Movie ) . To investigate whether CCV actin patches are needed for concentration of late endocytic vesicles , 3 dpi Vero cells were treated with latrunculin A ( LatA ) to depolymerize F-actin . Prior to LatA treatment , clusters of LAMP1+ vesicles were associated with F-actin patches punctuating the CCV . After LatA treatment , vesicles translocated to the juxta-nuclear region of the cell , leaving a large portion of the CCV devoid of LAMP1+ vesicles ( Fig 2B , S2 Movie ) . CCV CD63 was also redistributed although EEA1-labeled early endosomes were not ( S3A and S3B Fig ) . These results suggest actin patches serve as a scaffold that supports fusion of late endocytic vesicles with the CCV . To resolve whether actin polymerization is also needed for SNARE puncta , Vero cells at 3 dpi were treated with LatA , then stained for VAMP7 . Consistent with LatA-mediated detachment of late endocytic vesicles , VAMP7 was depleted on the CCV membrane ( Fig 2C and 2D ) . These data indicate actin polymerization promotes late endocytic SNARE clustering on the CCV membrane . The C . burnetti Dot/Icm type IVB secretion system ( T4BSS ) translocates effectors into the host cytoplasm that direct biogenesis of the CCV [17] . To test whether actin patch generation requires C . burnetii protein synthesis , Vero cells at 2 dpi were treated with chloramphenicol for 24 hr , then stained . Compared to untreated cells at 3 dpi , treated cells exhibited collapsed CCVs that lacked actin patches . Clustering of CD63+ vesicles on the CCV was also lost compared to untreated 3 dpi controls . The effects were reversible following washout of chloramphenicol and a 24 hr recovery period ( Fig 3A–3C ) . Similarly , chloramphenicol treatment reversibly decreased CCV clustering and colocalization of VAMP7 and Rab7 with actin ( S4A–S4D Fig ) . To confirm that CCV actin patch formation is T4BSS-dependent , Vero cells were infected with a C . burnetii dotA mutant and stained for F-actin and CD63 ( Fig 3D–3F ) [17] . At 1 dpi , the dotA mutant failed to induce formation of CCV F-actin patches associated with CD63+ vesicles . These results demonstrate that polymerization of CCV actin patches requires C . burnetii protein synthesis and Dot/Icm T4BSS effectors . Actin functions in various endosomal processes , such as endosome biogenesis , maturation , and transport [29 , 40–42] . Additionally , actin facilitates fusion between lysosomes and phagosomes [43 , 44] . Actin regulatory proteins control production of vesicular F-actin structures . Actin nucleators , such as annexin A2 and Arp2/3 , initiate actin filamentation , with Arp2/3 generating branched actin filaments . Arp2/3 is recruited and activated by NPFs , such as N-WASP and WASH . Moesin , an ezrin-radixin-moesin ( ERM ) family protein , and cortactin , bind and stabilize F-actin structures [40 , 41] . Ezrin , an ERM family protein , recruits N-WASP to endosomal membranes to initiate actin polymerization associated with phagosome-lysosome fusion [43] . Conversely , retromer recruits the WASH complex ( WASH , FAM21 , stumpellin , SWIP , CCD53 ) via FAM21 to endosomal membranes where it initiates actin polymerization that drives tubule scission and recycling of internalized membrane receptors from endosomes to the trans-Golgi ( retrograde trafficking ) or plasma membrane [31 , 33 , 42] . To identify regulators of actin polymerization at the CCV , Vero and THP-1 cells at 3 dpi were stained for vesicle-related actin regulators . VPS35 , FAM21 , WASH , Arp3 , and cortactin clustered with CCV actin patches , suggesting CCV actin polymerization is regulated by the retromer-WASH complex , with filaments stabilized by cortactin ( Fig 4A and 4B , S5 Fig ) . Arp2 phosphorylation , which is required for activation , was equivalent in infected and uninfected cells [45] ( S6 Fig ) . N-WASP exhibited minimal colocalization with CCV actin patches ( Fig 4A and 4B , S5 Fig ) . Ezrin , moesin , and annexin A2 did not localize to the CCV ( S7A and S7B Fig ) . To test whether the retromer-WASH complex association with CCV actin was directed by C . burnetii , Vero cells 2 dpi were treated with chloramphenicol for 24 hr , then fixed and stained for WASH or VPS35 ( S8A and S8B Fig ) . Corresponding to the loss of CCV actin patches following treatment , WASH and VPS35 clusters were also lost , an effect that was reversed following antibiotic washout . Furthermore , CCVs harboring a dotA mutant failed to recruit Arp3 ( S9 Fig ) . These data suggest C . burnetii effectors are required for recruitment and clustering of the retromer-WASH complex to mediate polymerization of CCV actin patches . Retromer sequesters membrane receptors for their retrieval from endosomes to the trans-Golgi or plasma membrane , a function that typically involves membrane tabulation and fission [31] . In infected cells , retromer also localizes with CCV actin patches associated with endosome fusion . To determine how retromer interacts with vesicles of the endosomal pathway during infection , Vero cells , left uninfected or infected for 3 days , were stained for retromer ( VPS35+ ) and early endosomes ( EEA1+ ) , late endosomes ( CD63+ ) , or lysosomes ( LAMP1+ ) . Consistent with previous reports [46 , 47] , VPS35 in uninfected cells associated with early and late endosomes , but minimally with lysosomes ( Fig 5A and 5C ) . In infected cells , colocalization of VPS35 with markers of late endosomes and lysosomes was increased ( Fig 5B and 5C ) , with an enrichment of label associated with the CCV . Thus , C . burnetii infection alters the cellular distribution of retromer . Sorting nexins ( SNXs ) are phosphatidylinositol 3-monophosphate ( PI3P ) binding proteins that coordinate with retromer to recycle receptors from endosomes . SNX1 or SNX2 dimerize with SNX5 or SNX6 and regulate canonical retrograde trafficking . SNX3 mediates an alternative route of cargo trafficking to the trans-Golgi in a WASH and actin-independent manner . SNX27 is involved in recycling plasma membrane receptors . [31 , 48] . In contrast to the near complete localization of VPS35 with CCV actin patches , SNXs moderately colocalized ( S10A and S10B Fig ) . To validate the specificity of antibodies directed against SNX1 and SNX2 , Vero cells infected for 1 day with Chlamydia trachomatis ( L2 serotype ) were immunostained . As previously reported [49] , SNX1 and SNX2 decorated the chlamydial inclusion ( S10C Fig ) . C . trachomatis and Legionella pneumophila inhibit retrograde trafficking via sequestration of SNX5/SNX6 and VPS29 , respectively , to the pathogen vacuole [50 , 51] . Because retromer and SNXs localized to the CCV membrane , we examined whether C . burnetii infection disrupts normal retrograde trafficking of CI-M6PR from endosomes to the trans-Golgi network . Inhibition of retrograde trafficking traps CI-M6PR in dispersed , early endosomes as opposed to a focused , juxta-nuclear localization within the trans-Golgi [47 , 52] . CI-M6PR dispersal did not occur upon knockdown ( KD ) by siRNA of VPS35 [53 , 54] ( Fig 6A , S11 Fig ) . Therefore , the chemical inhibitor Retro-2 was used which prevents vesicle fusion with the trans-Golgi network [55] . Without treatment , uninfected and infected Vero cells after 3 days of incubation showed focused CI-M6PR staining ( S12A and S12B Fig ) . In Retro-2 treated cells , CI-M6PR showed dispersed staining . Collectively , these data indicate infection does not disrupt retrograde recycling of CI-M6PR . Retromer recruits the WASH complex to endosomal membranes where it promotes actin polymerization required for scission of membrane tubules containing cargo for recycling [31 , 33 , 42] . To investigate whether retromer is necessary for CCV actin patch and vacuole formation , KD of VPS35 was performed ( Fig 6A ) . VPS35 KD inhibited CCV actin patch formation as witnessed by decreased actin patch size and intensity ( Fig 6B and 6C ) . Similar results were seen with KD of VPS29 in HEK 293 cells ( S13A–S13D Fig ) . VPS35 KD cells lacked FAM21 and WASH on the CCV membrane ( S14A–S14D Fig ) , indicating retromer is required for recruitment of the WASH complex to the CCV membrane and for efficient polymerization of CCV actin . Compared to cells treated with non-targeting ( NT ) siRNA , VPS35KD had no effect on C . burnetii replication but slightly decreased the size of the 3 day CCV ( Fig 6D and 6E ) . In addition to regulating fusion of late endosomes , Rab7 recruits retromer by binding VPS35 [56 , 57] . Focalized clusters of CD63 and Rab7 on the CCV in control cells were absent in VPS35 KD cells , where the markers localized uniformly around the CCV ( Fig 6B , S15A Fig ) . Diminished actin patches of VPS35 KD cells exhibited reduced colocalization with CD63 , suggesting CD63+ vesicles are not docking to CCVs at actin patches ( S15A and S15B Fig ) . Similarly , VAMP7 exhibited decreased clustering and colocalization with CCV actin patches compared to NT cells ( S15C and S15D Fig ) . Collectively , these results suggest retromer depletion allows Rab7 to uniformly distribute on the CCV , which promotes a similar distribution of late endocytic vesicles/fusion regulators . KD of VPS35 perturbed recycling of the glucose transporter 1 ( GLUT1 ) from endosomes to the plasma membrane , with trafficking redirected to lysosomes ( S16A Fig ) [58] . C . burnetii infection did not disrupt GLUT1 recycling to the plasma membrane . However , KD of VPS35 in infected cells dramatically increased the percentage of GLUT1+ CCVs ( S16B and S16C Fig ) . This indicates CCV growth is not dependent on retromer recycling of plasma membrane receptors . In VPS35 KD cells , the CCV lacks WASH and has reduced actin patches . To determine whether WASH is necessary for CCV actin patch formation and vacuole growth , we performed infections of inducible WASH knockout MEFs . Treatment with 4-hydroxy-tamoxifen ( 4-OHT ) of WASH ( f/f ) MEFs eliminates expression of WASH ( WASHout ) as determined by immunoblot and immunofluorescence ( S17A and S17B Fig ) . Similar to VPS35 KD cells , CCVs in WASHout cells at 3 dpi lacked actin patches and exhibited uniform distribution of Rab7 around the CCV ( Fig 7A and 7B ) . Rab7 redistribution was anticipated as retromer binds Rab7 [47] . CCV size and C . burnetii replication remained similar to untreated control cells ( WASHf/f ) ( Fig 7C and 7D ) . WASHout cells also displayed uniform staining of VPS35 around the CCV ( Fig 7E ) . Treatment of wild type MEFs ( WASH+/+ ) with 4-OHT did not affect actin patch formation , Rab7 clustering , CCV size or C . burnetii replication ( S18A–S18D Fig ) . Collectively , these results indicate that WASH depletion inhibits CCV actin patch formation , resulting in dispersion of retromer/WASH-generated actin sorting platforms . The Arp2/3 complex nucleates formation of barbed actin filaments involved in phagocytic uptake and endosomal trafficking [29] . Expansion and maintenance of the CCV is normal without WASH , an NPF that activates Arp2/3 , suggesting nucleation at the CCV is dispensable for C . burnetii growth [29] . To determine if Arp2/3-mediated actin polymerization in the endosomal pathway is necessary for CCV biogenesis , and to distinguish from phagocytic uptake of C . burnetii , Vero cells were treated at 1 dpi for 2 days with the specific Arp2/3 inhibitor CK-666 , which stabilizes the inactive state of Arp2/3 [59] . Like chloramphenicol-treated cells , CCVs in CK-666-treated cells lacked actin patches and failed to develop into spacious vacuoles supporting C . burnetii growth ( Fig 8A and 8B ) . Additionally , CCVs in Vero cells infected for 2 days , then treated for 1 day with CK-666 , displayed shrunken CCVs , an effect that was reversible ( Fig 8C and 8D ) . Loss of actin patches also correlated with loss of CCV Arp3 labeling ( S19A and S19B Fig ) . KD of Arp 3 also resulted in significantly smaller CCVs ( S19C , S19D and S19E Fig ) . Collectively , these results suggest CCVs of treated cells have a severe reduction in fusion with endosomes . To confirm inhibitory effects were not cell type specific , or related to retromer-WASH regulated trafficking or actin patch formation , WASHout cells at 2 dpi were treated for 1 day with CK-666 . Infected WASHout cells also had shrunken CCVs ( S20A and S20B Fig ) . This result indicated Arp2/3-mediated endocytic events upstream of vesicle fusion with CCVs are being disrupted . Fusion with late endosomes is required for CCV biogenesis [36] . Transferrin traffics to the CCV via the endosomal pathway [60] . Therefore , transferrin localization was examined in cells treated with CK-666 to determine if Arp2/3 is required for endosomal trafficking associated with CCV biogenesis . In Vero cells infected for 2 days , then treated with CK-666 for 24 hr , trafficking of 488-transferrin to CD63+ limiting membrane of CCVs was severely reduced ( Fig 9A–9C ) . These data demonstrate that Arp2/3-mediated actin dynamics that regulate trafficking within the endosomal pathway are required for CCV formation and pathogen growth .
Exploitation of the actin cytoskeleton is a common theme among intracellular bacteria [61] . Here , we show that F-actin patches form on the CCV membrane in Vero cells and THP-1 macrophages . Patches decorate with late endosome fusion proteins CD63 , Rab7 , VAMP8 , syntaxin 8 , and to a lesser extent , LAMP1 . Consistent with the presence of membrane fusion complexes , actin patches are preferred docking sites for vesicle fusion . Rapid disruption of CCV actin patches with LatA results in relocalization of LAMP1+ vesicles away from the CCV , without affecting vacuole size . Accordingly , this corresponds to the disappearance of CCV VAMP7 . Actin patches are less pronounced on non-expanding , mature vacuoles where fusion processes are likely down-regulated . Our data show that CCV actin patches also colocalize with retromer , Fam21 WASH , and Arp2/3 , and that the function of these proteins is necessary for patch formation . Several sorting nexins are also proximal to the retromer , although none strongly colocalize . SNX1 and SNX2 have previously been shown to localize in proximity of WASH and VPS29 [33 , 62] . Despite an association with CCV membrane fusion sites , retromer-generated actin is dispensable for CCV formation and C . burnetii growth . These conclusions are supported by robust replication of C . burnetii in VPS29 or VPS35 KD cells and WASH knockout MEFs , which lack CCV actin patches . Our results contrast with a previous study that reported individual silencing of VPS29 , VPS35 , SNX2 , SNX3 , SNX5 , or SNX6 inhibits C . burnetii infection . [11] . However , due to functional redundancy of SNX1 and SNX2 , as well as SNX5 and SNX6 , depletion of one of either pair will not disrupt retrograde trafficking [53 , 63] . Retromer is proposed to benefit C . burnetii by recycling beneficial host factors and/or removing harmful CCV determinants . Promoting proper CCV maturation was also suggested , although acidification of the CCV appears normal [11] . Disparate results may be due , in part , to different experimental readouts and cell lines ( e . g . HeLa versus HEK 239 cells ) . Unlike the neutral effect of retromer inhibition on C . burnetii growth , intracellular growth of Legionella pneumophila , C . trachomatis , and Salmonella typhimurium is negatively impacted by retrograde trafficking , and each pathogen secretes effector molecules to disrupt the process [49 , 51 , 64 , 65] . S . typhimurium inhibits retrograde trafficking of CI-M6PR through the activity of the type 3 secretion system ( T3SS ) effector SifA , which binds the host protein SifA- and kinesin-interacting protein ( SKIP ) . SifA-Skip sequesters Rab9 , which is necessary for late endosome to trans-Golgi transport of CI-M6PR . The end result is a Salmonella-containing vacuole with reduced lysosomal activity that enables pathogen growth [66] . The C . trachomatis T3SS effector IncE localizes to the chlamydial inclusion where it binds SNX5 and SNX6 [49 , 50 , 67] . Like S . typhimurium , subversion of retrograde trafficking is proposed to benefit chlamydial infection by disrupting lysosome function [50 , 67] . In L . pneumophila , the T4BSS effector RidL localizes to the Legionella-containing vacuole where it binds VPS29 and PI3P . These interactions restrict assembly of the retromer , thereby inhibiting retrograde trafficking [51] . The benefit to L . pneumophila is unclear . SNX1 , SNX2 , SNX5 , and SNX6 contain a Bin/Amphiphysin/Rvs ( BAR ) domain with curvature sensing , inducing , and stabilizing activities that promote endosomal tubulation associated with recycling of receptors from endosomes to the trans-Golgi ( retrograde transport ) or plasma membrane [48 , 63] . These SNXs , as well as SNX3 and SNX27 , contain a phox homology ( PX ) domain that preferentially binds PI3P on early endosomes . The canonical retromer involved in retrograde trafficking of cargo , such as CI-M6PR , is recruited to membranes by binding of VPS35 to both Rab7 and SNX3 [56 , 57] . The retromer involved in recycling plasma membrane receptors , such as GLUT1 , is recruited via binding of VPS26A or VPS26B to SNX27 [58 , 68–70] . Both of these retromer complexes utilize SNX-BAR and WASH-Arp2/3-mediated actin polymerization for tubule formation and scission [31] . The SNX3 retromer traffics cargo such as Wntless to the trans-Golgi where it is loaded with Wnt for transport to the plasma membrane [46 , 71 , 72] . The SNX3 retromer lacks the WASH complex and SNX-BARs , with cargo vesicles budding from the donor membrane through a process involving clathrin [46 , 56] . The appearance of GLUT1 on the CCV membrane following VPS35 KD is consistent with trafficking of the receptor to lysosomes for degradation following retromer disruption [58] . Thus , disruption of this recycling pathway , which controls trafficking of additional receptors , such as the monocarboxylate transporter [73] , is of little consequence to C . burnetii . Surprisingly , VPS35 KD did not disperse CI-M6PR to endosomes , suggesting retrograde trafficking is unaltered , and that delivery of hydrolases to the CCV occurs normally . In HeLa cells , a 6 . 5 h treatment with Retro-2 inhibits endosome-to-Golgi retrograde trafficking of toxins without affecting trafficking of endogenous cargos , like CI-M6PR [55] . The precise mechanism of inhibition is unknown but is speculated to involve direct inhibition or altered localization of Golgi syntaxin 5 . Based on a dispersed endosomal location of CI-M6PR , our results show trafficking of CI-M6PR in Vero cells is inhibited by Retro-2 following a 48 hr treatment . Retro-2 also did not inhibit CCV formation , which further supports the notion that canonical retrograde trafficking is dispensable for C . burnetii growth . This result also suggests C . burnetii does not require CI-M6PR-mediated delivery of hydrolases for growth . F-actin patches generated in a retromer-WASH complex and Arp2/3-dependent fashion are thought to generate force for scission of endosomal tubules [32 , 62] . VPS35-bound FAM21 recruits WASH to endosomal membranes [33 , 62] . On endosomes , retromer still binds in the absence of WASH , but is redistributed on the endosome membrane . We observed a similar redistribution of VPS35 on the CCV of WASH knockout MEFs . Activation of Arp2/3 by WASP family NPFs and subsequent generation of branched actin filaments modulates several membrane-associated functions including lamellipodia formation , receptor trafficking , endosome tubulation and shape , retrograde trafficking , autophagosome formation , and endocytosis/phagocytosis [30] . Accordingly , KD of Arp3 prior to infection of HeLa cells inhibits phagocytosis of C . burnetii [11] . Here , we show that treatment with the specific Arp2/3 inhibitor CK-666 , or Arp3 KD , inhibits CCV formation . Moreover , CK-666 treatment of cells harboring mature CCV causes vacuole collapse . Because actin filamentation mediates endocytic trafficking [29] , it is predicted to contribute to CCV maturation . Indeed , CK-666 inhibition of Arp2/3 inhibits endocytic trafficking of transferrin to the CCV . A previous study demonstrated C . burnetii protein synthesis is required for CCV recruitment of VAMP7 [36] . We show that chloramphenicol reversibly inhibits recruitment of VAMP7 , Rab7 , VSP35 , and WASH , as well as actin patch formation . Moreover , vacuoles harboring a dotA mutant lack actin patches . Collectively , these data suggest that recruitment of retromer and membrane fusion complexes is mediated by a C . burnetii T4BSS effector protein ( s ) . The CCV is unusual in being positive for both Rab7 , a marker of late endosomes/lysosomes , and enriched for PI3P , a characteristic of early endosomes [9] . Indeed , enrichment of PI3P on the CCV is associated with the activity of the C . burnetii effector CvpB . CvpB inhibits phosphotidylinositol 3-phosphate 5-kinase , the enzyme responsible for converting PI3P to phosphatidylinositol 3 , 5-bisphosphate on late endosomes [9] . Treatment with wortmannin , a phosphatidylinositol 3-kinase inhibitor , releases retromer and SNX1 from endosomes [47 , 63] . Thus , the possibility exists that retromer is aberrantly targeted to the CCV based on cooperative binding of VPS35 and SNX3 to Rab7 and PI3P , respectively , without serving a functional role [42 , 70] . The ability of CCVs devoid of actin patches , such as those in WASH knockout MEFs , to expand normally may be based on the dual function of Rab7 in membrane fission and fusion . In yeast , the Rab7 homolog Ypt7 is released upon membrane tubulation mediated by the SNX-BAR retromer , allowing interaction with the HOPS complex and membrane fusion [74] . Perhaps dispersal of Rab7 following depletion of VPS29 or VPS35 , or in WASH knockout cells , reflects the lack of F-actin organizing platforms that concentrate Rab7 at membrane fusion complexes [70] . Indeed , actin polymerization organizes WASH in discrete membrane domains [75] . Fusion of late endosomes/MVBs with the CCV would then occur in a dispersed manner , as opposed to localized fusion associated with actin patches . The ability of C . burnetii to prosper without CCV actin may also be related to the hyper-fusogenic nature of the vacuole [76] . During the preparation of this manuscript , two controversial studies were coordinately published indicating SNX-BAR without retromer can mediate retrograde retrieval of CI-M6PR [53 , 54] . Data derived by quantitative proteomics and protein binding assays show CI-M6PR directly binds SNX5 and SNX6 [53 , 54] . Moreover , retrograde retrieval of CI-M6PR is unaffected by KD or knockout of retromer components [53 , 54] . This behavior is associated with SNX-BAR and retromer occupying separate microdomains on endosomes , as suggested by a previous report [77 , 78] . As with our investigation , these studies employed the broadly used and accepted assay for dysfunctional retrograde sorting , i . e . , dispersal of CI-M6PR into endosomes . Others have suggested that disparate results arise from different measures of CI-M6PR dispersal [79] . Moreover , a two-step model is proposed that invokes a role for retromer in transiently concentrating cargo , which is then rapidly transferred to SNX-BAR for tubule-mediated retrieval to the Golgi [53 , 77] . Clearly , more study is needed , although our data showing poor colocalization of SNXs with retromer , and the lack of CI-M6PR dispersal following VSP35 KD , agrees with the newly proposed model . Nonetheless , plasma membrane recycling of GLUT1 is clearly inhibited upon retromer disruption to result in aberrant trafficking of the receptor to the CCV membrane . Based on the bulk of published data , we propose a model , similar to one recently proposed by Jimenez-Orgaz et al . [80] , that invokes a dual regulatory role for CCV Rab7 in endosome fusion and retromer recruitment ( S21 Fig ) . Membrane receptors , such as GLUT1 , are recycled from transitioning early to late endosomes by retromer . This can occur coincidently with late endosome engagement with the CCV , but prior to vesicle fusion . Retromer recruitment by CCV Rab7 , and subsequent WASH-Arp2/3 generation of actin sorting platforms , concentrates Rab7 and impedes its lateral diffusion . Consequently , recruitment of HOPs/SNARE complexes by Rab7 results in focused fusion of late endosomes with the CCV . In the absence of retromer and sorting platforms , membrane receptors are not recycled but instead deposit on the CCV in a dispersed manner due to the unfocused nature of Rab7 . While retromer-derived F-actin is dispensable for C . burnetii growth , additional Arp2/3-generated F-actin structures are essential for CCV biogenesis and pathogen replication , including those that mediate endocytic trafficking . An intriguing question is the identity and function of TBSS effector ( s ) modulating this fusogenic behavior . In summary , this study illustrates the complex interplay between the CCV and actin-mediated vesicular trafficking pathways . It also describes a large vacuole model for studying retromer function , reminiscent of vacuoles generated by ectopic expression of constitutively-activated Rab5 [54] .
HEK 293 ( ATCC CRL-1573 , human embryonic kidney epithelial ) cells were cultured in Delbecco’s Modified Eagle Medium ( DMEM ) ( Life Technologies ) supplemented with 10% fetal bovine serum ( FBS ) . THP-1 ( ATCC TIB-202 , human monocytic ) and Vero ( ATCC CCL-81 , African green monkey epithelial ) cells were cultured in RPMI medium 1640 ( Life Technologies ) supplemented with 10% FBS . WASH conditional knockout mouse embryonic fibroblasts ( MEFs ) ( Daniel Billadeau , Mayo Clinic ) were grown in DMEM containing 10% FBS . Generation and characterization of these cells are described in Gomez et al . [81] . MEFs were grown in DMEM containing 10% FBS , and knockout of WASH was obtained with two sequential 24 hr treatments of 3 μM 4-OHT ( Sigma , H7904 ) . After treatment , MEFs were washed and cultured for an additional 6–7 days before use in experiments [81] . All cell lines were incubated at 37°C with 5% CO2 . Infection of Vero cells with C . trachomatis was at a multiplicity of infection ( MOI ) of 50 based on inclusion forming units . C . burnetii Nine Mile phase II , RSA439 ( NMII ) was propagated in Vero cells and purified as described [82] . The C . burnetii dotA mutant was propagated in the synthetic medium ACCM-2 as described [17] . C . trachomatis ( LGV-434 , serotype L2 ) was propagated in HeLa cells and purified as described [83] . For infection , HEK 293 , Vero , or MEF cells were seeded on coverslips in 24-well plates at a density of 4 x 104 cells per well . THP-1 cells were seeded on coverslips in 24-well plates at a density of 3 x 105 cells per well and stimulated with 200 nM phorbol myristate acetate ( Sigma ) for 24 hr for differentiation into macrophage-like cells and attachment to coverslips . For immunofluorescence staining , cells were infected at an MOI of 100 based on genome equivalents ( GE ) quantified by TaqMan qPCR using a StepOnePlus Real-Time PCR system ( Applied Biosystems ) and primers specific to C . burnetii groEL [38] . For C . burnetii growth analysis , cells were infected at an MOI of 10 . Latrunculin A ( LatA; Sigma ) was used at 1 μg/ml , CK-666 ( Sigma ) at 200 μM , chloramphenicol at 50 μg/ml , and Retro-2 ( Sigma ) at 40 μM . All treatments were at 37°C with 5% CO2 in RPMI media 1640 ( Life Technologies ) plus 10% FBS . DMSO was used as a control treatment as necessary . Vero cells on glass bottoms of 24-well SensoPlates ( Greiner bio-one , 662892 ) were transfected with CellLight Actin-RFP , BacMam 2 . 0 ( Molecular Probes , C10502 ) 1 dpi and CellLight Lysosomes-GFP , BacMam 2 . 0 ( Molecular Probes , C10507 ) the night before imaging at 3 dpi . CellLight reagents were used at concentrations recommended by the manufacturer . Cells were washed to remove CellLight reagents before imaging on a Nikon ECLIPSE Ti spinning disk confocal fluorescence microscope . For imaging of LAMP1+ vesicles fusing with CCV membranes , cells were incubated at 37°C with 5% CO2 and imaged at 1 frame every 2 sec . For LatA ( Sigma ) treatment , cells were imaged at room temperature at 1 frame per min . Imaging started 15 min before adding LatA followed by 30 min post-treatment . Cells were fixed with 4% paraformaldehyde in phosphate-buffered saline ( PBS ) for 30 min at room temperature and simultaneously permeabilized and blocked for 30 min with 0 . 1% Triton X-100 or 0 . 05% saponin plus 1% BSA . Antibodies ( 3–5 μg/ml ) were diluted in Triton or saponin buffers and samples stained for 30–60 min . The following antibodies were used for immunofluorescence and/or immunoblotting: Actin ( Abcam , ab8224 ) , Annexin A2 ( clone: D11G2 , cell signaling , 8235 ) , Arp2 ( Abcam , AB128934 ) , phosphoArp2 ( Abcam , AB119766 ) , Arp3 ( Millipore , 07–272 ) , CD63 ( clone: H5C6 , BD Pharmingen , 556019 ) , Cortactin ( clone: 30/Cortactin , BD Transduction Laboratories , 610049 ) , EEA1 ( Cell Signaling , 2411 ) , EEA1 ( clone: 14/EEA1 , BD Transduction Laboratories , 610456 ) , Ezrin ( Cell Signaling , 3145 ) , FAM21C ( Millipore , ABT79 ) , GLUT1 ( Abcam , ab15309 ) , LAMP1 ( Abcam , ab24170 ) , LAMP1 ( clone: 1D4B , Santa Cruz , sc-19992 ) , Moesin ( clone:EP1863Y , Abcam , ab52490 ) , Moesin ( clone: 38/Moesin , BD Transduction Laboratories , 610401 ) , Rab7 ( clone: D95F2 , Cell Signaling , 9367 ) , SNX1 ( clone: 51/SNX1 , BD Transduction , 611482 ) , SNX2 ( clone: 13/SNX2 , BD Transduction , 611308 ) , SNX3 ( Abcam , ab56078 ) , SNX27 ( clone: 1C6 , Abcam , ab77799 ) , Syntaxin 8 ( clone: 48/syntaxin 8 , BD Transduction , 611352 ) , TNG46 ( Sigma , T7576 ) , VAMP7 ( US Biological , V2024 ) , VAMP8 ( Synaptic System , 104302 ) , VPS29 ( Sigma , HPA039748 ) , and VPS35 ( Abcam , ab10099 ) . Rabbit anti-WASH VCA domain-specific antibodies were kindly provided by Dr . Daniel Billadeau [81] . Rabbit anti-human WASH and guinea pig anti-human N-WASP antibodies were kind gifts from Dr . Matthew Welch [29] . Anti-C . burnetii antibodies were generated in guinea pigs or rabbits , and anti-C . trachomatis in rabbits . Alexa Fluor-647 , 568 , and 488-conjugated secondary antibodies ( Life Technologies ) were used . For staining F-actin , BODIPY 558/568 or Alexa Fluor-647 labeled phalloidin ( Life Technologies ) were used . Nuclei were stained with Hoescht 33342 ( ThermoFisher ) . Post-staining , cells were again fixed with 4% paraformaldehyde for 30 min and coverslips mounted using Prolong Gold antifade mountant ( ThermoFisher ) . Dharmacon ON-TARGETplus SMARTpool siRNAs for VPS29 ( L-009764-01-0005 ) , VPS35 ( L-010894-00-0005 ) , Arp3 ( L-012077-00-0005 ) , and non-targeting ( D-001810-10-5 ) were used . HEK 293 cells were used as opposed to HeLa cells because more complete and consistent KD was obtained . For KDs , HEK 293 cells were grown on coverslips in 24-well plates at a density of 4 x 104 cells per well for 1 day before transfection . Dharmafect 1 ( Dharmacon ) was used to transfect cells with siRNA complexes at a concentration of 2 μM . For KD of VPS29 and VPS35 , cells were infected two days post-transfection . For KD of Arp3 , cells were transfected 1 and 3 days before infection . Cells were fixed 3 dpi and processed for immunofluorescence . Immunoblots of lysates were used to confirm efficiency of KD . Actin served as a loading control . MEF or HEK 293 cells seeded at 4 x 104 cells per well in 24-well plates were infected with C . burnetii at an MOI of 10 by centrifuging at 500 x g for 30 min at room temperature . Cells were then washed and incubated with growth medium . Samples were collected by trypsinization , bead beaten with 0 . 1 mm Zirconia/silica beads ( BioSpec ) in a homogenizer ( ThermoElectron FastPrep FP120 ) , and boiled 10 min . GE were quantified by qPCR . CI-M6PR retrograde trafficking assays were performed as described in Osborne et al . [52] . HEK 293 cells were incubated with 10 μg/ml anti-CI-M6PR ( BioRad , MCA2048 ) in serum-free medium for 1 hr at 37°C . Cells were washed with PBS prior to immunostaining . Vero cells were treated with DMSO or Retro-2 for 2 days , followed by incubation for 1 hr at 37°C with 10 μg/ml anti-CI-M6PR in serum free medium . Cells were washed with PBS , then fixed and fluorescently stained . Vero cells seeded on glass coverslips at 4 x 104 cells per well in 24-well plates were infected at an MOI of 100 . Two dpi cells were treated with DMSO or CK-666 for 24 hr . Cells were incubated with 25 μg/ml of Alexa Fluor-488 transferrin ( ThermoFisher , T13342 ) for 1 hr , washed three times with PBS , fixed with 4% PFA , then fluorescently immunostained for CD63 and C . burnetii . Fixed and stained cells were imaged using Zeiss LSM-710 confocal fluorescence microscope ( Carl Zeiss ) . For compiling images , z-sections of 0 . 32 μm slices were taken . Depicted images of cells are maximum intensity projections of 3 slices , totaling 1 μm thickness . Fiji ( Image J , National Institutes of Health ) was used for all image analysis . The entire CCV was selected for co-localization and fluorescence intensity analysis and included a combined measurement of 3 z-sections from the vertical center ( z-axis ) of the CCV . The area of CCVs was measured using CD63 , LAMP1 , or Rab7 as CCV membrane markers . For determining disruption of CI-M6PR retrograde trafficking , the percentage of cells with visually compact versus scattered CI-M6PR staining was quantitated . Correlation coefficients ( Pearson’s Correlation Coefficient ) were determined using ImageJ Coloc 2 . Unless otherwise stated , a minimum of 60 cells for each condition from 3 independent experiments was used for analyses . GraphPad Prism ( GraphPad Software ) was used for all graphing and statistics .
|
Coxiella burnetii , the human Q fever bacterium , replicates in a harsh , lysosome-like compartment termed the Coxiella-containing vacuole ( CCV ) . Here , the pathogen directs formation of the CCV through the activities of secreted type 4B ( T4B ) effector proteins . Questions remain concerning how the host cytoskeleton and receptor recycling pathways contribute to creation and function of the CCV . We found that filamentous actin ( F-actin ) patches formed around the CCV in a T4B-dependent manner . Patches were preferred sites of endosome-CCV fusion , a behavior that correlated with patch enrichment of endosome fusion proteins , such as VAMP7 and Rab7 . Patch formation required colocalized retromer-WASH-Arp2/3 recycling complexes . In cells depleted of retromer or WASH , CCV lacked actin patches and displayed Rab7 , also involved in retromer recruitment , uniformly redistributed around the CCV membrane . C . burnetii grew normally in these cells , indicating retromer-mediated protein recycling and CCV actin patches are dispensable for productive infection . In contrast , global disruption of Arp2/3-generated F-actin severely restricted CCV formation through a process associated with defective endosome trafficking . We propose that retromer sorting and formation of CCV actin patches are inconsequential to pathogen growth . These studies refine our understanding of vesicular trafficking pathways required for CCV biogenesis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"vero",
"cells",
"medicine",
"and",
"health",
"sciences",
"vesicles",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"biological",
"cultures",
"microbiology",
"membrane",
"fusion",
"cellular",
"structures",
"and",
"organelles",
"bacterial",
"pathogens",
"endosomes",
"research",
"and",
"analysis",
"methods",
"coxiella",
"burnetii",
"contractile",
"proteins",
"specimen",
"preparation",
"and",
"treatment",
"actins",
"staining",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"cell",
"lines",
"cell",
"membranes",
"biochemistry",
"cytoskeletal",
"proteins",
"cell",
"staining",
"cell",
"biology",
"biology",
"and",
"life",
"sciences"
] |
2018
|
Actin polymerization in the endosomal pathway, but not on the Coxiella-containing vacuole, is essential for pathogen growth
|
Lymphatic endothelial cells ( LECs ) are differentiated from blood vascular endothelial cells ( BECs ) during embryogenesis and this physiological cell fate specification is controlled by PROX1 , the master regulator for lymphatic development . When Kaposi sarcoma herpes virus ( KSHV ) infects host cells , it activates the otherwise silenced embryonic endothelial differentiation program and reprograms their cell fates . Interestingly , previous studies demonstrated that KSHV drives BECs to acquire a partial lymphatic phenotype by upregulating PROX1 ( forward reprogramming ) , but stimulates LECs to regain some BEC-signature genes by downregulating PROX1 ( reverse reprogramming ) . Despite the significance of this KSHV-induced bidirectional cell fate reprogramming in KS pathogenesis , its underlying molecular mechanism remains undefined . Here , we report that IL3 receptor alpha ( IL3Rα ) and NOTCH play integral roles in the host cell type-specific regulation of PROX1 by KSHV . In BECs , KSHV upregulates IL3Rα and phosphorylates STAT5 , which binds and activates the PROX1 promoter . In LECs , however , PROX1 was rather downregulated by KSHV-induced NOTCH signal via HEY1 , which binds and represses the PROX1 promoter . Moreover , PROX1 was found to be required to maintain HEY1 expression in LECs , establishing a reciprocal regulation between PROX1 and HEY1 . Upon co-activation of IL3Rα and NOTCH , PROX1 was upregulated in BECs , but downregulated in LECs . Together , our study provides the molecular mechanism underlying the cell type-specific endothelial fate reprogramming by KSHV .
Kaposi sarcoma ( KS ) was originally described by a Hungarian doctor Morris Kaposi in 1872 , but did not attract much scientific , clinical and public attention until 1980s when acquired immunodeficiency syndrome ( AIDS ) became an epidemic and KS was subsequently found to be the most common cancer among HIV-positive individuals [1] . KS-associated herpes virus ( KSHV ) , also known as human herpes virus ( HHV ) -8 , was identified in 1994 as the causative agent for KS [2] . KSHV is a member of the gamma herpes virus superfamily and , similar to other herpes virus , has distinct latent and lytic stages . While KS tumor formation is initiated by latent infection of host cells by KSHV , a small fraction of latent cells spontaneously undertakes the lytic phase , a reproductive replication process that releases the infectious viral particles for another round of infection [1] . During the latent infection , KSHV expresses only a handful of genes among its some 90 viral genes and these latent genes include latency-associated nuclear antigen ( LANA ) , viral cyclin-D homolog ( vCyc-D ) , viral FLICE-inhibitory protein ( v-FLIP ) and kaposin isoforms [3]–[8] . KS tumors are often associated with vessel-like spaces that are filled with immune cells and red blood cells , and the proliferating tumor cells of KS are now believed to be originated from KSHV-infected vascular endothelial cells [9] . The endothelial origin of the KS tumor cells was first proposed about 45 years ago based on the expression of blood vascular endothelial cell ( BEC ) markers by KS cells [10] . However , identification of new signature genes for lymphatic endothelial cells ( LECs ) in the 1990s prompted a re-evaluation on the histogenetic origin of the KS tumor cells . A number of research groups have subsequently reported the expression of LEC-markers in all stages of KS tumors , in addition to the previously identified BEC-specific molecules [11] . Moreover , KSHV has been reported to be able to infect both BECs and LECs under in vitro culture condition [12] . Importantly , we and other have reported that KSHV induces a host cell fate reprogramming [13]–[15] . When KSHV infects BECs , the virus upregulates a significant number of LEC-signature genes . On the contrary , when KSHV infects LECs , the host cells express some of the BEC-associated genes . Importantly , neither of these host cell fate reprogramming processes is a complete trans-differentiation , but rather KSHV infection directs the host cells to move away from their original cell identities and to end up somewhere in between the two endothelial cell fates , exhibiting mixed cell phenotypes , based on the genome-wide transcriptional profiling studies [12]–[15] . Thus , considering the fact that BECs differentiate to become LECs during embryogenesis [16] , KSHV-infected BECs are considered to undertake a “forward” differentiation , and KSHV-infected LECs go through a “reverse” differentiation [12]–[15] . We and others have demonstrated that KSHV induces this host cell type-specific reprogramming by deregulating the expression of PROX1 [12]–[15] , a homeodomain transcriptional regulator that functions as the master control protein in LEC-differentiation [16] , [17] . Whereas KSHV upregulates PROX1 in BECs where PROX1 is not expressed , the virus downregulates PROX1 in LECs , a cell type which abundantly expresses PROX1 [12]–[15] . This host-specific bidirectional regulation of PROX1 and host cell fate reprogramming by KSHV are of great biological and pathological significance to better understand the host-virus interaction . Nonetheless , the underlying molecular mechanism remains undefined . Several signal transduction pathways and transcriptional factors have been identified to regulate the expression of PROX1 [18]–[22] . Among them , interleukin ( IL ) -3 and NOTCH signals have been previously implicated with KS pathogenesis [23]–[26] . The IL-3 receptor ( IL3R ) consists of a heterodimer of an α-subunit and a β-subunit: While IL3R α-subunit ( IL3Rα/CD123 ) determines the receptor specificity for IL3 , IL3R β-subunit ( IL3Rβ/CD123 ) , having a long cytoplasmic domain , is responsible for transducing external signals and can form heterodimers with IL5 receptor α-subunit and granulocyte-macrophage colony stimulating factor ( GM-CSF ) receptor α-subunit to make IL5R and GM-CSFR , respectively [27] . IL-3 , a potent stimulator of differentiation of hematopoietic stem cells , was found to trans-differentiate cultured BECs to LECs by upregulating PROX1 expression [18] . Moreover , IL-3 is constitutively expressed and secreted by cultured LECs and required to maintain the LEC-phenotype in vitro [18] . On the other hand , we have reported that NOTCH signal suppresses the lymphatic phenotype by downregulating lymphatic cell fate regulators , PROX1 and COUP-TFII , through its effector Hey1 and Hey2 [22] . In this report , we defined the molecular mechanism underlying the host-specific bidirectional regulation of PROX1 by KSHV . We found that KSHV activates both IL3Rα and NOTCH pathways and that these two pathways opposingly regulate PROX1 expression in blood versus lymphatic-lineage endothelial cells . Our data show that IL3Rα pathway serves as an activating signal for PROX1 expression in KSHV-infected BECs and human umbilical vein endothelial cells ( HUVECs ) , whereas NOTCH acts as a repression signal for PROX1 expression in KSHV-infected LECs . In summary , KSHV-induced co-activation of these two opposing signals results in PROX1-upregulation in KSHV-infected BECs and HUVECs , but PROX1-downregulation in KSHV-infected LECs . Together , our data provide the molecular mechanism underlying the cell type-specific regulation of PROX1 by KSHV .
We and others have previously reported that KSHV infection of primary human BECs resulted in upregulation of otherwise silenced PROX1 along with a lymphatic reprogramming of host cell fate [13]–[15] . Interestingly , however , it was also reported that KSHV infection of primary human LECs resulted in downregulation of PROX1 [12] . To investigate whether these seemingly inconsistent results may be due to experimental variations among different research groups , or due to endothelial lineage ( BEC vs . LEC ) -specific differential regulation of PROX1 by KSHV , we performed a series of comparative analyses of KSHV-mediated PROX1 regulation by using the same donor-derived primary neonatal human BECs and LECs , along with HUVECs from different donors . In agreement with the previous reports [12]–[15] , quantitative real time RT-PCR ( qRT-PCR ) revealed that KSHV-infection upregulated PROX1 in BECs and HUVECs , while downregulating PROX1 in LECs ( Fig . 1A ) . Western blot analyses further confirmed that the steady-state level of PROX1 protein was increased in BECs and HUVECs by KSHV infection , but significantly decreased in LECs upon KSHV-infection . Thus , our study confirmed that KSHV differentially regulates the expression of PROX1 in blood versus lymphatic-lineage endothelial cells by upregulating PROX1 in BECs and HUVECs where PROX1 is not expressed , while downregulating in LECs where PROX1 is abundantly expressed . Previous studies have demonstrated that IL-3 activates PROX1 expression in BECs and HUVECs , but not in LECs , [18] and that KSHV infection of microvascular endothelial cells strongly induces expression and secretion of IL-3 [26] . These reports have led us to investigate whether IL-3 signaling plays a role in PROX1-upregulation by KSHV in BECs and HUVECs . Indeed , KSHV-infection significantly upregulated the expression of IL3 receptor alpha ( IL3Rα/CD123 ) mRNA and protein in BECs , HUVECs and LECs ( Fig . 1B ) . We also determined the expression of related receptors such as IL3Rβ , IL5Rα and GM-CSFRα . While IL3Rβ , the other subunit of IL3 receptor , was found to be upregulated by 4-fold in KSHV-infected BECs , IL5Rα and GM-CSFRα , which can dimerize with IL3Rβ to form receptors for IL5 and GM-CSF , respectively , were not regulated by KSHV ( Fig . 1C ) . Moreover , immunohistochemical analyses of KS tumor sections demonstrated prominent expression of IL3Rα in human dermal KS tumor cells ( Fig . 1D ) . Because IL-3 activates PROX1 expression in BECs and HUVECs , but not in LECs [18] , we investigated whether overexpression of its receptor , IL3Rα , could upregulate PROX1 in each cell type . We found that adenoviral expression of IL3Rα resulted in a strong upregulation of PROX1 mRNA and protein in BECs and HUVECs , but not in LECs ( Fig . 2A ) , suggesting that the PROX1-upregulating signal by IL3Rα is operative only in PROX1-deficient BECs and HUVECs and does not affect the already abundant expression of PROX1 in LECs . Furthermore , ectopic expression of IL3Rα in BECs and HUVECs resulted in upregulation of various LEC-signature genes such as podoplanin ( PDPN ) , VEGFR3 , FGFR3 , LYVE1 , CDKN1C ( p57Kip2 ) , ITGA1 ( integrin α1 ) , PPL ( periplakin ) and SLC ( secondary lymphoid chemokine ) in both BECs and HUVECs ( Fig . 2B ) , indicating that IL3Rα may play an important role in lymphatic reprogramming of KSHV-infected BECs and HUVECs . We next set out to determine whether IL3Rα is essential for KSHV-mediated PROX1 upregulation in BECs and HUVECs , and thus inhibited IL3Rα using siRNAs or an IL3Rα-neutralizing antibody . Indeed , both approaches showed that PROX1 inhibition could efficiently abrogate the KSHV-induced upregulation of PROX1 mRNA and protein ( Fig . 2C ) . These data demonstrate that IL3Rα plays an integral role in PROX1 upregulation by KSHV in BECs and HUVECs , but not in LECs . The key downstream mediators for the IL3/IL3Rα pathway include Jak2 and STAT5a/b , and activated STAT5a/b proteins rapidly enter the nuclei and bind to the promoters of target genes to modulate their gene expressions [28] . We thus investigated whether STAT5a/b proteins are involved in the IL3Rα-mediated PROX1 upregulation . Adenoviral overexpression of IL3Rα in BECs , HUVECs and LECs revealed an increased phosphorylation in STAT5a/b in BECs and HUVECs , but at much lesser degree in LECs ( Fig . 3A ) . We then overexpressed the wild type , constitutive active or dominant negative form of STAT5b protein in BECs , HUVECs and LECs by transient transfection and determined their effects on PROX1 expression . Interestingly , the ectopic expression of the constitutive active form of STAT5b protein resulted in upregulation of PROX1 protein in BECs and HUVECs , but did not change PROX1 expression in LECs ( Fig . 3B ) . Moreover , the IL3Rα-induced PROX1 upregulation was significantly abrogated when the expression of STAT5a/b was inhibited by siRNA-mediated knockdown ( Fig . 3C ) , establishing a key role of STAT5a/b in the IL3Rα-induced PROX1 upregulation . To further corroborate the molecular interaction between STAT5a/b protein and the PROX1 gene , we next searched for putative binding sites of STAT5a/b proteins in the PROX1 promoter regions of different animal species and mapped two or three candidate sites in the PROX1 promoter regions of nine mammals ( human , chimp , mouse , rat , guinea pig , horse , rabbit , dog and marmoset ) ( Supplemental Fig . S1 ) . Importantly , both DNA sequences and relative locations of the putative STAT5a/b binding sites were found to be highly conserved among the nine mammalian species . To further validate them as STAT5a/b binding sites , we next performed chromatin immunoprecipitation ( ChIP ) and gel electrophoretic mobility shift assays ( EMSA ) for the human PROX1 promoter region . Indeed , our ChIP assay detected a basal binding activity of endogenous STAT5 proteins to the two putative STAT5a/b sites ( termed E1 and E2 ) found approximately 11 . 7- and 6 . 5-kb upstream , respectively , from the human PROX1 initiation codon in BECs and HUVECs , and these binding activities were significantly increased by IL3Rα overexpression ( Fig . 3D ) . Moreover , EMSA demonstrated that nuclear extracts isolated from IL3Rα-overexpressed BECs efficiently caused a shift in the mobility of both E1 and E2 probes ( Fig . 3E ) . Addition of an anti-STAT5a/b antibody in the EMSA reactions significantly inhibited formation of the protein/probe complexes ( Fig . 3F ) , indicating that E1 and E2 probes made DNA/protein complexes with STAT5a/b proteins . Taken together , these data demonstrate that STAT5a/b proteins play a significant role in the IL3Rα-induced PROX1 upregulation in BECs and HUVECs by directly binding to the PROX1 promoter region . We next set out to investigate how KSHV-infection resulted in downregulation of PROX1 in LECs , despite the fact that KSHV upregulates PROX1 in BECs and HUVECs . Previous studies have shown an increased activity of the NOTCH pathway in KSHV-infected endothelial cells and KS-tumor cells in vivo [23] , [25] , [29] , [30] . Moreover , we have recently reported that activated NOTCH represses PROX1 expression through HEY1 in LECs [22] . Accordingly , we came up with a hypothesis that KSHV-induced NOTCH activation may be involved in PROX1 downregulation in KSHV-infected LECs . Supporting this hypothesis , KSHV-infection of all three cell types , BECs , HUVECs and LECs , resulted in upregulation of HEY1 ( Fig . 4A ) . Adenoviral overexpression of NOTCH intracellular domain ( NICD ) caused a significant downregulation of PROX1 in LECs , but not in BECs and HUVECs ( Fig . 4B&C ) . In addition , Notch activation in LECs resulted in downregulation of additional lymphatic-signature genes such as podoplanin ( PDPN ) and CDKN1C , suggesting a suppressive role of Notch signaling in LEC phenotypes ( Supplemental Fig . S2 ) . Moreover , microarray-based analyses on the NICD-induced modulation of the transcriptional profiles in primary LECs ( National Center for Biotechnology Information , Gene Expression Omnibus accession number: GSE20978 ) support the effect of Notch on LEC phenotype . Furthermore , we found that HEY1 , like NICD , was able to strongly repress the expression of PROX1 protein , when overexpressed in LECs ( Fig . 4D ) . Since HEY1 is known to repress target gene expression by binding to the promoter [31] , we performed HEY1-ChIP assays against the PROX1 promoter in primary LECs and found that HEY1 was indeed physically associated with the PROX1 promoter around the two transcriptional start sites ( Fig . 4E&F ) . We then generated a set of PROX1-promoter reporter constructs and found that a 1 . 8-kb proximal promoter region was sufficient to deliver the HEY1-mediated repression ( Fig . 4G ) . Finally , inhibition of HEY1 expression by siRNA abrogated the KSHV-mediated downregulation of PROX1 mRNA and protein in LECs ( Fig . 4H ) . Together , these findings demonstrate that NOTCH activation is responsible for the KSHV-mediated PROX1 downregulation in LECs , but not in BECs and HUVECs , and that the NOTCH effector HEY1 directly binds to the PROX1 promoter to downregulate its gene expression . We have previously reported that PROX1 physically and functionally interacts with the orphan nuclear receptor COUP-TFII to specify the cell fate of LECs [32] . We also performed a genome-wide search for PROX1 target genes using microarray analyses and identified a list of genes , whose expression was altered by PROX1 knockdown in LECs [32] . Interestingly , the microarray analyses revealed that the expression of HEY1 was significantly downregulated in LECs by PROX1 knockdown ( GEO accession: GSE12846 ) . This unexpected regulation of HEY1 by PROX1 was further confirmed using qRT-PCR of total RNAs isolated from LECs that were transfected with siRNA against PROX1 and/or COUP-TFII ( Fig . 5A ) . Notably , knockdown of COUP-TFII , a PROX1-interacting protein , did not alter the HEY1 expression . On the contrary , adenoviral overexpression of PROX1 , but not COUP-TFII , in LECs resulted in a strong upregulation of HEY1 ( Fig . 5B ) . We then asked whether PROX1 could activate the proximal promoter of HEY1 and thus performed a series of luciferase reporter assays using promoter constructs of HEY1 and two other HEY family members , HEY2 and HEYL . Indeed , the HEY1 promoter was found to be activated by PROX1 wild type , but not by a PROX1 mutant lacking DNA-binding activity [33] ( Fig . 5C ) . Moreover , adenoviral expression of PROX1 in human umbilical aortic endothelial cells ( HUAEC ) also resulted in upregulation of HEY1 ( Fig . 5D ) . We then performed PROX1 ChIP assays against the HEY1 promoter and found that PROX1 protein is physically associated with the HEY1 promoter ( Fig . 5E ) . Subsequently , a set of HEY1 promoter reporter constructs was generated and used to further study the PROX1 regulation of HEY1 . Notably , a ∼0 . 7 kb-long HEY1 promoter ( pHey1C ) was sufficient to deliver the PROX1-mediated activation of the HEY1 promoter ( Fig . 5F ) . Therefore , we concluded that PROX1 positively regulates the expression of HEY1 by directly binding to its promoter . Together with the findings above ( Fig . 4 ) , these data established a reciprocal regulation between PROX1 and HEY1: HEY1 functions as a repressor of PROX1 and PROX1 is required to upregulate or maintain HEY1 expression . According to this reciprocal feedback regulation , PROX1 could negatively regulate its own gene expression . To confirm this auto-regulation , we ectopically overexpressed PROX1 in LECs using adenovirus that harbors the PROX1 open reading frame ( ORF ) only and then determined the expression level of the endogenous PROX1 by two qRT-PCR probes detecting the PROX1 3′-untranslated region ( UTR ) , which is not present in the adenovirus . Indeed , ectopic expression of PROX1 resulted in a significant downregulation of the endogenous PROX1 ( Fig . 5G ) . Taken together , our data uncovered an intricate auto-regulatory mechanism for the PROX1 gene expression that utilizes the HEY1 repressor , a component of NOTCH signal pathway . Since our studies above showed that KSHV activates both IL3Rα and NOTCH pathways simultaneously , we next asked how these two signals counteract with each other in regulating the expression of PROX1 in BECs versus LECs . To address this question , we concurrently activated both pathways by adenoviral expression of IL3Rα and NICD in BECs , HUVECs and LECs . Notably , co-expression of IL3Rα and NICD resulted in differential regulation of the PROX1 expression in blood vs . lymphatic-lineage endothelial cells ( Fig . 6A , B ) . In BECs and HUVECs , PROX1 was found to be upregulated by co-expression of IL3Rα and NICD . In LECs , however , PROX1 was rather downregulated by activation of the two pathways . These data indicate that the IL3Rα-induced PROX1 activating signal is more effective than the NICD-mediated PROX1 repression in BECs and HUVEC , resulting in PROX1 upregulation . On the contrary , the NICD-mediated repression is more prominent than the IL3Rα-induced activation in LECs , causing downregulation of PROX1 . Taken together , co-activation of the IL3Rα and NOTCH pathways yields a differential expression of PROX1 and may account for the KSHV-mediated endothelial lineage-specific differential regulation of PROX1 and accompanying host cell fate reprogramming .
Despite their morphological and functional similarities , endothelial cells exhibit remarkable heterogeneity and plasticity . Heterogeneity of endothelial cells is profoundly contributed by their plastic cell fates in response to internal and external stimuli such as immunological , functional , metabolic , anatomical and hemodynamic signals [34] , [35] . In addition to these physiological stimuli , pathological insults can also incite the plasticity of endothelial cell identities . We and others have previously reported that KSHV-infection reprograms the host cell identity by triggering a drift from their original phenotypes and that PROX1 , whose expression is deregulated by KSHV , plays a key role in this pathological host cell fate reprogramming [12]–[15] . Although this concept of the pathogen-induced host cell fate reprogramming has forwarded a new view on the histogenetic origin of KS tumor cells [12]–[15] , two subsequent questions , how and why , remained to be answered . In this study , we aimed to address the first question , how , by studying the KSHV-mediated regulation of PROX1 in context of the virus-induced host cell fate plasticity . By using primary BECs and LECs from the same donors , we confirmed that KSHV-infection induces endothelial cell type-specific differential regulation PROX1 , as previously reported [13]–[15] . Because only unidirectional differentiation ( BECs to LECs ) occurs during physiological ( embryonic ) endothelial cell differentiation , it is important to understand how KSHV pathologically activates a differentiation program by upregulating PROX1 in one cell type and a de-differentiation by downregulating PROX1 in another cell type [12] . Based on the data presented above , we propose a novel hypothesis for the molecular mechanism underlying the KSHV-mediated host cell type-specific regulation of PROX1 ( Fig . 7A ) . We hypothesize that KSHV-infection results in simultaneous activation of the IL3Rα and NOTCH pathways and delivers both positive and negative regulatory signals , respectively , to the PROX1 gene in both blood and lymphatic-lineage endothelial cells . Under this condition , we believe that it is the initial expression status of PROX1 that brings up the differential consequences: An activating signal ( IL3Rα ) will cause a discrete change than a repressive signal ( NOTCH ) in the PROX1-deficient BECs and HUVECs , whereas a repressive signal ( NOTCH ) will have a higher impact than an activating signal ( IL3Rα ) in the PROX1-expressing LECs . Moreover , because LECs constitutively secret IL-3 to maintain their phenotypes [18] , IL3Rα-mediated signaling is already active in LECs and thus the increased expression of IL3Rα by KSHV does not provide an additional upregulation of PROX1 . Accordingly , KSHV-mediated upregulation of PROX1 directs a “forward” reprogramming ( differentiation ) of BECs and HUVECs to acquire the lymphatic phenotype , but KSHV-mediated downregulation of PROX1 enables a “reverse” reprogramming ( dedifferentiation ) of LECs . Importantly , since both pathological cell fate reprogramming are incomplete processes , KSHV-infection forces host endothelial cells to move away from their original cell fates and to end up somewhere in between the two endothelial cell fates , as described by previous studies [12]–[15] . Another key finding in the current study is the reciprocal regulation between PROX1 and HEY1 ( Fig . 7B ) . We found that HEY1 , the only transcriptional repressor of PROX1 identified to date [22] , directly binds to the PROX1 proximal promoter to repress its transcription . Moreover , our current study revealed that PROX1 is required to maintain the expression of HEY1 in LECs . This PROX1/HEY1 reciprocal regulation put forward three important implications . First , a positive regulatory signal of PROX1 expression will be always counteracted by the parallel upregulation of its own repressor HEY1 and thus PROX1 expression will be maintained under a certain threshold level that is set by HEY1 . This speculation is consistent with the finding that the activating signal by IL3Rα did not additionally increase PROX1 expression in LECs where PROX1 is already abundantly expressed ( Fig . 2A ) . Second , since HEY1 has been known to repress the expression of the viral lytic phase initiator gene RTA and thus to play an important role in the latency control [36] , [37] , PROX1 may indirectly contribute to host regulation of the viral latency phase by maintaining the HEY1 expression . Third , HEY1 requires PROX1 for its own expression and thus will be unable to entirely shut down PROX1 expression , even if NOTCH signal is activated . This regulation will serve as a feedback control to counteract the NOTCH-induced repression of PROX1 , which is mediated by HEY1 . Therefore , the reciprocal regulation between PROX1 and HEY1 adds another layer of complexity to the opposing regulatory circuits of PROX1 expression by IL3Rα and NOTCH upon KSHV-infection . Moreover , since KS cells secrete a number of chemokines and cytokines and also recruit various immune cells , multiple signal transduction pathways have been associated with KS tumorigenesis [8] , [9] . Considering the numerous chemokines and cytokines that KS cells are constantly exposed to [26] , [38] , [39] , it is likely that PROX1 expression would be controlled by multiple activating and repressing signals . In fact , activation of Jak2/Stat3 by gp130 has been demonstrated to be important in KSHV-mediated upregulation of PROX1 and lymphatic reprogramming [40] . Gp130/Stat3 and IL3Rα/STAT5 may cooperate to regulate the PROX1 expression in KSHV-infected cells , especially when the two pathway are known to cross-talk with each other [41] . Supporting this notion , inhibition of IL3Rα in KSHV-infected BECs and HUVECs only partially abrogated PROX1 upregulation by KSHV ( Fig . 2C ) . One important question brought up by the current study is to determine which KSHV viral proteins are responsible for the upregulation of IL3Rα and NOTCH . We speculate that the latent protein kaposin-B may be involved in upregulation of IL3Rα because kaposin-B has been reported to upregulate expression of various cytokines such as granulocyte-macrophage colony-stimulating factor ( GM-CSF ) via stabilizing their mRNAs [42] and , notably , GM-CSF has been demonstrated to be a key activator of IL3Rα expression in different cell types [43]–[46] . Therefore , it will be very interesting to investigate if kaposin-B regulates IL3Rα expression . On the other hand , the molecular mechanism underlying the KSHV-mediated NOTCH upregulation has been recently documented by two studies [25] , [47] , which identified the KSHV-encoded vFLIP , LANA and vGPCR to be responsible for the upregulation of various NOTCH signal components , including the receptors ( NOTCH1∼4 ) , ligands ( Dll1/4 , Jag1 ) and downstream effectors ( Hey1 ) of the NOTCH signaling pathway . Although our current study addresses the how part to a certain extent , numerous questions on why remain unanswered: Why does KSHV induces its host cell fate reprogramming ? It will be intriguing to determine whether this host cell fate reprogramming is a mere by-product of the viral infection of endothelium , a cell type that happens to be highly plastic in their cell identity , or whether the compromised cell identity provides any pathological advantage to KS tumorigenesis . Further studies should be warranted on these important questions .
Human primary dermal blood vascular endothelial cells ( BECs ) and lymphatic endothelial cells ( LECs ) were isolated from de-identified neonatal human foreskins and cultured as previously described [32] with an approval of the University of Southern California Internal Review Board ( PI: YK Hong ) . Primary human umbilical venous endothelial cells ( HUVECs ) and human umbilical aortic endothelial cells ( HUAEC ) were purchased from Lonza ( Basel , Switzerland ) and cultured in EGM-2 medium ( Lonza ) . Primary endothelial cells were transfected with siRNA by electroporation as previously described [48] . Sequence information of siRNAs are as follows: human HEY1 [22] , PROX1 [22] , STAT5a/b ( SC29495 , Santa Cruz Biotechnology ) , COUP-TFII ( UCGUACCUGUCCGGAUAUA , UAUAUCCGGACAGGUACGA ) , control ( fire fly luciferase , CUUACGCUGAGUACUUCGAdTdT ) , and IL3Rα ( CUGGGACCUUAACAGAAAUdTdT ) . Adenovirus for PROX1 [49] , COUP-TFII [32] and NICD [50] were previously described . Adenovirus expressing V5-tagged NICD was a kind gift from Dr . Lucy Liaw ( Maine Medical Center Research Institute ) [51] . Adenovirus for human IL3Rα was constructed by transferring the IL3Rα coding sequences from AxCALNLhIL-3Rα ( Riken BRC DAN Bank , Japan ) into Ad-Track-CMV shuttle vector and then recombined with AdEasy-1 based on the reported protocol of adenovirus construction [52] . Expression vectors for Flag-tagged rat-STAT5b ( wild type , constitutive active , dominant negative ) [53] were kind gifts from Dr . Peter Rotwein ( Oregon Health and Science University ) . KS specimens were provided from the AIDS and Cancer Specimen Resource ( ACSR ) with an approval of the University of Southern California Internal Review Board ( PI: YK Hong ) . Infectious KSHV was purified from BCBL-1 cells as we previously described [54] . Endothelial cells were infected with KSHV for 2∼4 days and infectivity was measured by western blot analyses for KSHV LANA . For inhibition of IL3Rα , BECs or HUVECs were transfected with IL3Rα siRNA or treated with anti-IL3Rα antibody ( 1 µg/ml ) 24 hours before KSHV-infection and 48 hours post infection , RNA and whole cell lysates were harvested for further analyses . Blocking of STAT5a/b expression by siRNA was similarly performed . Real-time RT-PCR was performed by using TaqMan EZ RT-PCR Core Reagent ( Applied Biosystems ) . Each reaction was multiplexed for target gene and β-actin for normalization . Sequences of two Taqman primer/probes for the 3′ UTR are ( TGGTTTTCCCTTTTACAATCGAA/GAATTTGGAGAGACAGGCTTTTG/FAM-TTGTGCCTCCCAAGTGCATTGGAA-TAMRA; TGGTTTTCCCTTTTACAATCGAA/GAATTTGGAGAGACAGGCTTTTG/FAM-TTGTGCCTCCCAAGTGCATTGGAA-TAMRA ) . Sequences for other primers and probes used for this study will be provided upon request . Immunostaining was performed on formalin-fixed paraffin embedded KS tissue specimens by following a standard immunostaining protocol [32] . Sources of antibodies for immunostaining or western blot analyses are PROX1 ( Millipore Corporation , MA ) , β-actin and Flag ( Sigma-Aldrich Corporation ) , STAT5 and phospho-STAT5 ( Cell Signaling Technology ) , IL3Rα ( clone 7G3 , BD Bioscience ) , V5 ( Invitrogen ) and LANA ( Advanced Biotechnologies Inc , Maryland ) . ChIP assays were performed as previously described [22] . BECs , HUVECs or LECs transduced with a control or IL3Rα-expressing adenovirus for 48 hours were subjected to ChIP assays by using anti-STAT5a/b antibody . Genomic/protein precipitants were PCR-amplified by using primers against the E1 site ( CTTCCCTTCTTCAGGGTGCT/TCACGCCTCCTGTTCTTTCT ) or the E2 site ( TAGCTCAAGGAGGCAGGTTG/GGGCATGAGTGGAAAAGAGA ) sites ( Fig . 4A ) . Sequences of the primers used for HEY1 ChIP assays against human PROX1 are as follows ( Primer Set 1: GAGAGGCTCGGTCCCACT/TGAGTAATGGGAGGCTCTTTTC; Primer Set 2: GAGCCTCCCATTACTCAGACC/GAGGCTCCCGCTTAGAAACT ) . EMSA was performed as previously described [22] , [54] . Briefly , nuclear lysates isolated from BECs or LECs transduced with a control or IL3Rα-expressing adenovirus were incubated with 32P-labeld oligonucleotide probes containing the putative E1 site for STAT5a/b ( ATCTGGTTGTAATTCTCAGAATTGGTT/TCCTAAACAAACCAATTCTGAGAATTA ) or the E2 site ( GCTTGTTTTTATTTTTCCGAGAAGATC/GACAGCACAGATCTTCTCGGAAAAATA ) in the binding buffer and subjected to polyacrylamide gel electrophoresis . For EMSA blocking assays , the nuclear extracts were added to the binding reactions in the presence of a STAT5a/b antibody ( 1 µg/ml ) . The reporter constructs of PROX1 promoter were constructed as follows . A 5 . 1-kb human PROX1 promoter was PCR-amplified from human genomic DNA using primers ( GTCCAGGGCGTGTACTGAG/CGGCTGCAATGGTGTATTATT ) and cloned into EcoRV site of pBluescriptIISK ( - ) in a reverse orientation . A NheI/XhoI fragment was then transferred to NheI/XhoI sites of pGL3 ( Promega ) to generate the PROX1_P1 vector . Subsequently , a 1 . 6-kb PstI or a 3 . 3-kb MluI fragment was deleted from PROX1_P1 and self-ligated after Klenow treatment to construct PROX1_P2 and PROX1_P3 , respectively . Luciferase reporter constructs [55] for mouse HEY1 , HEY2 and HEYL were generously provided by Dr . Eric Olson ( University of Texas Southwestern Medical Center , Dallas ) . The reporter constructs of the mouse Hey1 promoter were constructed by modifying a 2 . 9-kb mouse Hey1 construct [56] ( renamed here as pHey1A ) generously provided by Dr . Manfred Gessler ( Theodor-Boveri-Institut fuer Biowissenschaften , Germany ) . Deletion constructs were generated by self-ligation after digesting pHey1A with EcoRV/MscI ( pHey1B ) , EcoRV/ApaI ( pHey1C ) , EcoRV/AscI ( pHey1D ) , EcoRV/SacII ( pHey1E ) or EcoRV/NcoI ( pHey1F ) . Luciferase assays were performed as follows . HEK293T was transiently transfected in 12-well plates in DMEM , 10% FCS . Cells were transfected with a total of 1 ug of DNA by Lipofectamin 2000 ( Invitrogen ) . After 48 hours , cells were washed and lysed in 200 µL of PBS by three-time-repetition of freeze-thaw cycles . Cells were harvested and cell debris was removed by centrifugation . Protein concentrations were measured by Bradford assay and luciferase activity was measured using the Bright-Glo buffer ( Promega ) by Mikrowin2000 program , Luminometer ( Plate CHAMELE , HIDEX ) . Each experiment was repeated 3 times with each reaction measured in triplicates .
|
Kaposi's sarcoma ( KS ) is one of the most common neoplasms in HIV-positive individuals and organ transplant recipients . KS-associated herpes virus ( KSHV ) , also known as human herpes virus ( HHV ) -8 , has been identified as the causative agent and infects endothelial cells to form KS . Importantly , we and others have discovered that when KSHV infects endothelial cells of blood vessels , it reprograms host cells to resemble endothelial cells in lymphatic vessels . On the other hand , when KSHV infects endothelial cells in lymphatic vessels , the virus directs the host cells to partially obtain the phenotypes of blood vessel endothelial cells . These host cell reprogramming represent abnormal pathological processes , which are not as complete as the physiological process occurring during embryonic development . Currently , it is not clear how and why this cancer causing virus modifies the fate of its host cells . In this study , we aimed to dissect the molecular mechanism underlying the virus-induced host cell fate reprogramming and found two important cellular signaling pathways , interleukin-3 and Notch , playing key roles in the pathological events . Our current study provides a better understanding of KS tumorigenesis with a potential implication in a new KS therapy .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"mechanisms",
"of",
"signal",
"transduction",
"endothelial",
"cells",
"microbiology",
"host-pathogen",
"interaction",
"cancers",
"and",
"neoplasms",
"feeback",
"regulation",
"dna",
"transcription",
"oncology",
"bone",
"and",
"soft",
"tissue",
"sarcomas",
"gene",
"expression",
"biology",
"kaposi",
"sarcoma",
"molecular",
"biology",
"signal",
"transduction",
"cell",
"biology",
"virology",
"cellular",
"types",
"molecular",
"cell",
"biology"
] |
2012
|
Opposing Regulation of PROX1 by Interleukin-3 Receptor and NOTCH Directs Differential Host Cell Fate Reprogramming by Kaposi Sarcoma Herpes Virus
|
As model organism-based research shifts from forward to reverse genetics approaches , largely due to the ease of genome editing technology , a low frequency of abnormal phenotypes is being observed in lines with mutations predicted to lead to deleterious effects on the encoded protein . In zebrafish , this low frequency is in part explained by compensation by genes of redundant or similar function , often resulting from the additional round of teleost-specific whole genome duplication within vertebrates . Here we offer additional explanations for the low frequency of mutant phenotypes . We analyzed mRNA processing in seven zebrafish lines with mutations expected to disrupt gene function , generated by CRISPR/Cas9 or ENU mutagenesis methods . Five of the seven lines showed evidence of altered mRNA processing: one through a skipped exon that did not lead to a frame shift , one through nonsense-associated splicing that did not lead to a frame shift , and three through the use of cryptic splice sites . These results highlight the need for a methodical analysis of the mRNA produced in mutant lines before making conclusions or embarking on studies that assume loss of function as a result of a given genomic change . Furthermore , recognition of the types of adaptations that can occur may inform the strategies of mutant generation .
The recent increased use of reverse genetic approaches has been largely driven by the ease , affordability of construction , and implementation of the CRISPR/Cas9 and TALEN systems . Recent communications recount numerous cases of generated mutations in genes of interest lacking an expected effect on phenotypes [1 , 2] . The shift from antisense-based knockdown ( morpholinos , RNAi ) to mutant generation ( gene targeting/TILLING methods ) resulted in discrepancies in phenotypes , leading researchers to question the specificity and mechanisms of anti-sense technologies and also the methods by which mutants are generated [3] . A screen for essential genes performed in a human cultured cell line found little correlation between genes identified with short hairpin RNA ( shRNA ) silencing and CRISPR/Cas9 methods [4] . While genome editing methods , such as the CRISPR/Cas9 and TALEN systems , have proven to be an efficient and effective way to reduce or eliminate gene function , a frequent lack of a mutant phenotype is observed , often explained by genetic compensation . This is a process wherein related genes or pathway members are differentially regulated in the mutants to compensate for the targeted loss of a specific gene [3] . In addition to genetic compensation , other mechanisms to recover the function of genes harboring homozygous mutations involve variations in processing of mRNA . For example , variations in essential splice sites ( ESS ) in humans often lead to loss of function resulting in disease [5 , 6]; however , there are several well described ways that function may be recovered [7–9] . In canonical pre-mRNA splicing , joining exons for a functional product requires the presence of a 5’ splice donor sequence ( intronic GU ) , a branchpoint adenosine , the polypyrimidine tract , and a splice acceptor sequence ( intronic AG ) . Base variations in the ESSs lead to one of four outcomes , in order of frequency: 1 ) exon skipping , 2 ) activation of cryptic splice sites , 3 ) activation of cryptic start sites producing a pseudo-exon within the intron or 4 ) intron inclusion , in the case of short or terminal introns [10] . Mutations in ESSs that lead to skipped exons may result in transcripts that escape nonsense-mediated decay ( NMD ) , the surveillance system that reduces errors in gene expression , if the exon skip does not lead to a frame shift and premature translation termination codon ( PTC ) [11] . Cryptic splice sites are present throughout the genome both by chance and through evolution of introns [12] and their activation and use by splicing machinery is typical when exon definitions ( such as the natural splice sequences ) have been altered [13 , 14] . Depending on the location of the cryptic splice site used , and the impact on the sequence and frame , functional transcripts may still be generated . Nonsense-associated alternative splicing , in which a PTC-containing exon is skipped , may also restore the reading frame of a mutated gene [8] . Again , if the exon skip does not lead to a frame shift and new PTC , and the skipped exon does not contain essential motifs , transcripts may be generated that escape nonsense-mediate decay . Location of the PTC also determines whether the nascent transcripts will be subject to NMD [15]; however , even though these transcripts escape the surveillance system that detects PTCs , these transcripts may either be functional or aberrant . Translation of the transcripts may result in wildtype or deleterious function . Recently , discussions on how to produce successful knockout models have been renewed [16–18] . To better inform the generation of future mutations , this report analyzes the genetic consequences of several chemical- and CRISPR-induced zebrafish mutant lines in depth . Zebrafish are amenable to current genome editing methods [19 , 20] and are a well-established vertebrate model routinely used to assign functions to genes through the use of classical genetic approaches [21] . To begin to investigate the type and frequency of adaptations that may lead to unexpected splicing in the context of mutation , we carried out studies of mutant lines that included measuring transcript levels and analyzing mRNA splicing ( cDNA sequence ) and genomic sequence for the presence and use of cryptic splice or start sites . Of the seven mutant lines presented in this study , we show five examples that result in altered mRNA processing . Our findings emphasize the need to analyze putative mutant lines at the level of the mRNA sequence and not assume that a mutation will have the predicted effect on mRNA and/or loss of function .
To determine whether each ZMP line containing an ESS mutation results in the predicted skipping of an exon , adult heterozygote mutant zebrafish were incrossed and their offspring were pooled or individually processed into total RNA and cDNA . PCR amplification of this cDNA template revealed amplicons of the expected size from each ZMP line , while abca1bsa18382 ( ATP-binding cassette transporter , sub-family A , member 1B ) and slc27a2asa30701 ( solute carrier family 27 , member 2a; protein is fatty acid transport protein , member 2a ) also had a shorter amplicon ( 116 and 210 bp shorter , respectively; S1 Fig ) that matches the predicted length of amplicons of these cDNAs after the omission of the affected exon . The pla2g12bsa659 ( phospholipase A2 , group XIIB ) mutant allele has a mutation in the essential splice acceptor site preceding its final ( fourth ) exon and could not be investigated for the loss of that exon using these methods . To confirm that exons are skipped in abca1bsa18382 and slc27a2asa30701 and determine why the mutations in abca1asa9624 ( ATP-binding cassette transporter , sub-family A , member 1A ) and cd36sa9388 ( cluster of differentiation 36 , aka fatty acid translocase ) did not appear to lead to the predicted skipping of exons ( S2 Fig ) , individual 6-dpf larvae underwent genotyping , gDNA and cDNA sequence analysis , and qPCR studies . abca1bsa18382 has a point mutation in the essential splice acceptor site of intron 33–34 ( g . 64427G>T ) ( Fig 1 ) . To determine whether the point mutation results in skipping of the subsequent exon ( e34 ) and use of the ( next ) essential splice acceptor site of intron 34–35 , we performed PCR amplification using primers targeted to flanking exons of cDNA ( synthesized from individual , genotyped larvae; 713-bp amplicon ) , followed by Sanger sequencing . As expected , cDNA sequencing confirmed that exon 34 ( 116 bases ) is skipped and leads to a frame shift in abca1bsa18382/+ and abca1bsa18382/sa18382 larvae . Following the frame shift , the mutant cDNA encodes a 13 AA open reading frame ( ORF ) and an early termination signal that would direct the loss of exons 34–46 . qPCR studies reveal transcript levels are down 3 . 5-fold in 6-dpf abca1bsa18382/sa18382 zebrafish ( ANOVA with Tukey’s test , p = 0 . 049 ) . slc27a2asa30701 has a point mutation in the essential splice donor site of intron 2–3 ( g . 3431G>A ) ( Fig 2 ) . cDNA sequencing confirms omission of exon 2 in slc27a2asa30701/+ and slc27a2asa30701/sa30701 larvae . No frame shift is observed since exon 2 is 210 bases long ( encoding 70 AA ) . By qPCR , transcript levels in 6-dpf slc27a2asa30701/sa30701 zebrafish did not differ from those of their wildtype siblings ( ANOVA with Tukey’s test; p-value greater than threshold of 0 . 05 ) . abca1asa9624 has a point mutation in the 3’ ESS of intron 29–30 ( g . 48320G>A ) ( Fig 3 ) . Analysis of cDNA sequence from individual genotyped larvae revealed the loss of three bases , “TAG” , at the start of exon 30 in heterozygous and homozygous mutants . To look for cryptic splice sites , a flanking region of gDNA was PCR amplified and sequenced . A cryptic splice acceptor site ( “AG” ) was found 2 and 3 bases downstream of the mutated wildtype splice acceptor site , in exon 30 . Use of this cryptic splice acceptor site splices out the first three bases of exon 3 ( “TAG” ) and the protein this message encodes would lack one Serine ( and remain in frame with the wildtype product ) . Transcript levels in 6-dpf abca1asa9624/sa9624 zebrafish did not differ from their wildtype siblings ( ANOVA with Tukey’s test; p-value>0 . 05 ) . cd36sa9388 has a point mutation in the 5’ ESS ( splice donor site ) of intron 10–11 ( g . 11242G>A ) ( Fig 4 ) . cDNA sequencing of individual , genotyped larvae reveals incorporation of extra bases “ATAT” in between the sequence for exon 10 and exon 11 , which leads to a frame shift in the mutant allele . After the frame shift , 18 AA and a PTC follow , predicting the loss of exon 12 ( 154 AA ) . The PTC position sits at the last exon-exon junction and thus transcripts are predicted to escape NMD . Transcript levels of 6-dpf cd36sa9388/sa9388 larvae did not differ significantly from their wildtype siblings ( ANOVA with Tukey’s test; p-value>0 . 05 ) . To look for the use of a cryptic splice donor site , a flanking region of gDNA isolated from individual larvae was amplified and sequenced . The wildtype sequence at the 5’ end of intron 10–11 includes the splice donor “GT” . However , in the mutant allele , the first base is mutated to an “A” , resulting in the loss of the splice donor site . The mutated intronic sequence begins “ATATGT…” , which provides a cryptic splice donor site ( “GT” ) 3 and 4 bases downstream of the mutated wildtype splice donor site ( Fig 4 ) . pla2g12bsa659 has a point mutation in the splice acceptor site of intron 3–4 ( g . 10194A>T ) and is predicted to skip the last ( 4 of 4 ) exon; thus , exons flanking the mutation could not be PCR amplified to confirm the loss of exon 4 in mutants . Attempts to amplify an alternative transcript with retention of either the final exon 4 or the intron 3–4 did not succeed when using cDNA synthesized from homozygous mutant larvae as the template . During phenotypic screening and genotyping of 5-dpf larvae from heterozygous incrosses , a total of 29 pla2g12bsa659/sa659 larvae exhibited a darkened yolk phenotype while 52 pla2g12b+/+ siblings did not ( 2 experiments; S3 Fig ) . Correspondingly , RNA expression profiling demonstrates a 3-fold decrease in mutants/siblings ( adj . p = 3 . 68 x 10−8 ) [23] ( S2 Table ) . creb3l3asa18218 has a nonsense mutation in exon 2 of 10 ( g . 357C>T ) , which changes codon CAA to TAA , a PTC ( Fig 5 ) . PCR amplification of cDNA ( wildtype and homozygous pooled larval intestines ) followed by gel electrophoresis revealed two bands in homozygous mutants but only the expected wildtype band in the wildtype siblings ( Fig 5 ) . cDNA sequencing of the bands showed alternative transcripts with the unexpected omission of exon 2 in homozygous mutant but not in wildtype larvae . Splicing out exon 2 ( 114bp encoding 38 AA ) does not lead to a frame shift . The nonsense mutation was found to occur in a predicted exonic splice enhancer ( ESE ) sequence using the web-based prediction tool , ESEFinder [24] . Mutation of an ESE , an important aspect of exon definition , could explain a reduction of transcripts that include exon 2 in mutant cDNA . By qPCR , transcript levels of 6-dpf creb3l3asa18218/sa18218 dissected intestines did not differ significantly from their wildtype siblings ( Wilcoxon-Mann-Whitney test; p>0 . 05 ) . It has been shown that conserved alternative exons have a high percentage of preservation of reading frame [11] and 41% of all human exons are symmetrical ( divisible by 3 ) [25] . To determine whether zebrafish exons 2–10 were symmetrical at a higher frequency than expected by random chance , we summarized the remainder ( 0 , 1 , and 2 ) for all coding genes in Ensembl GRCz10 ( Release 90; August 2017 ) across each exon . Our analysis of zebrafish coding exons 2–10 revealed between a 5 . 1% and 7 . 2% increase over chance ( 33 . 33% ) in exons divisible by 3 ( Chi-squared test , df = 2 , p-value < 2 . 2e-16 ) ; S4 Fig ) . To confirm that these adaptive phenomena are not specific to ENU-mutagenized lines , we analyzed a 7-bp deletion in exon 3 of smyd1a ( g . 6948_6955del; SET and MYND domain containing 1A ) which was generated using CRISPR/Cas9 targeting methods . The 7-bp deletion leads to a predicted frame shift and PTC ( 48/485 AA produced ) ( Fig 6 ) . To look for evidence of novel alternative splicing , smyd1a cDNA was sequenced from wildtype and mutant embryos by cloning full-length PCR products . As expected , all 20 wildtype cDNA clones had the smyd1a wildtype sequence and all 20 clones from the homozygous mutant embryos contained the 7-bp deletion in exon 3 . However , 6 of the 20 cDNA clones from mutant embryos exhibited alternative splicing at exon 2 ( Fig 6 ) . Three clones had an alternative splice event using a cryptic splice acceptor site ( “AG” ) in exon 2 , located 13-bp downstream of the wildtype splice acceptor site , leading to a 13-bp deletion at the 5’ end of the exon 2 . Similarly , sequence data from another three clones show the use of a cryptic splice acceptor site ( “AG” ) 40 bp downstream of the wildtype splicing site , resulting in a 40-bp deletion at the 5’ region of exon 2 . Both deletions are predicted to lead to a frame shift and premature translation termination . qPCR studies revealed transcript levels of 1- and 2-dpf smyd1amb4/mb4 zebrafish were down 13-fold compared to wildtype siblings ( Wilcoxon-Mann-Whitney test , p = 0 . 000077 ) .
In this report , we analyzed the compensatory mechanisms that function through permissive mRNA processing in the context of ENU- and CRISPR-induced mutations ( Table 2 ) . Recently , Popp et al . reviewed how the process of exon-junction-complex-mediated NMD influences the success of creating loss-of-function mutations with CRISPR/Cas9 [17] . Most notable is their earlier finding that NMD cannot occur if a PTC is within 50–55 nucleotides ( nt ) of the last exon-exon junction [15] . In our study , we found one example of this phenomenon . For the cd36sa9388 allele , the resultant PTC is within 1 nt of the last exon junction ( e11–e12 ) and as predicted , we observed wildtype transcript levels in the homozygous mutants . Others have proposed identifying potential cryptic start sites before the construction of any CRISPR or TALEN vectors , after finding wildtype expression levels in in vitro mouse NIH3T3 cell lines harboring frame-shift mutations in Gli3 [16] . Loss of function from mutations near the translation initiation site may be recovered by utilizing nearby downstream alternative translation initiation sites [26] . The mutations in our lines were closer to the middle or 3’ end of genes . We did identify use of cryptic splice sites in the mutant allele in three of seven lines ( abca1asa9624 , cd36sa9388 , smyd1amb4 ) , underlining the importance of identifying potential cryptic splice sites prior to basing studies on presumed lack of gene function . We also described an example of nonsense-associated splicing ( creb3l3asa18218 ) , wherein a PTC-containing exon is spliced out and creb3l3asa18218/sa18218 larvae have wildtype transcript levels . The mechanisms underlying this process are still being explored: in many cases , mutations in conserved splice elements ( such as exonic splice enhancers; ESE ) have been shown to cause nonsense-associated splicing [27–31] . Prykhozhij et al . also recently illustrated the need for careful mutation analysis , beyond the level of gDNA sequence . They found only one of three mutant zebrafish lines resulted in the predicted frameshift [18 , 32] . Of the remaining two lines , one displayed an exon skip , possibly due to a mutation in an ESE , and the other used an alternative start site . Moreover , there have been two recent publications documenting numerous cases of exon skipping in response to CRISPR/Cas9-mediated mutations [33 , 34] . Our analysis of whether zebrafish coding exons 2–10 are divisible by 3 greater than 33 . 33% of the time revealed a 5 . 1–7 . 2% increase over expected . Taken together , these data suggest exon skipping in response to mutation is more common than generally thought and support our suggestion that when possible , researchers target exons that are not divisible by 3 . Since ESS mutations often lead to human disease [6] , in vivo models are critical to our understanding . However , we found that skipping an exon may still lead to a viable product: if the exon is divisible by 3 and thus its omission does not lead to a frame shift and PTC , transcript levels were not subject to NMD ( creb3l3asa18218 , slc27a2a sa30701 ) . In both lines found to skip an exon in our study , sequence alignment with their human ortholog revealed no known essential motifs in the skipped exons ( S5 Fig ) . While the skipped exon 2 in zebrafish contains part of the ATP/AMP binding motif responsible for fatty acid activation ( through acyl-CoA synthetase activity ) , data from functional studies suggest that it functions efficiently in long-chain fatty acid transport through the FATP/VLACS motif [35] . Of the two human splice isoforms , FATP2a and FATP2b , the latter lacks the ATP/AMP binding motif but has the FATP/VLACS motif . Expressing FATP2b in yeast and mammalian cultured cells revealed that it functions in long chain fatty acid transport . Examination of intron-spanning reads from available temporal expression data revealed no evidence of the alternative transcripts we identified in this study in wildtype larvae [36] , suggesting that they did not result from wildtype alternative splicing events . These data are not consistent with a low-abundance mRNA variant that is normally expressed in the WT background emerging to partially compensate for the loss of the major WT mRNA variant in the mutant background . In this study , we report that five of seven analyzed zebrafish lines with induced mutations show evidence of compensation through altered mRNA processing and contribute to the growing data of how to produce successful knockout models . Our data support a hypothesis that there may be a surveying mechanism that could detect mutations and adapt mRNA alternative splicing to cope with potential loss of function . Our findings are consistent with an analysis of 418 nonsense gene variants in the human population that catalog very similar adaptations and suggests “that permissive RNA processing and translation in human cells facilitates the accumulation of otherwise deleterious genetic variation in the human population” [37] . Analysis of cDNA sequence in mutant alleles may allow for prediction of compensation , simply by scanning for proximal cryptic splice and initiation sites that might be used for alternative transcripts . Moreover , it is entirely possible that splice-blocking morpholinos could engage some of the same compensatory mechanisms described in this study . This hypothesis can be tested by future studies in which cDNA from morphants are subjected to sequencing and supports our contention that researchers always sequence the cDNA in mutants and morphants . Employing multiple “guide” RNAs in the CRISPR/CAS9 system can result in large intron-spanning deletions in or the elimination of targeted genes . While this approach has been used to generate loss-of-function alleles , it can lead to the deletion of the genomic regions needed for post-transcriptional regulation of gene expression or transcriptional regulation of other genes . It is estimated that 30–80% of human coding genes are post-transcriptionally regulated , at least in part , by microRNAs ( miRNAs ) [38]; so far , 2 , 619 miRNAs and 324 , 219 miRNA-target interactions have been annotated in human ( miRTarBase ) [39] and approximately 40% of miRNA genes are located within the introns of protein-coding genes [40] . Rather than creating large deletions or removing an entire gene , other approaches , such as those used to generate nonsense mutations or small deletions , may work better to generate loss-of-function alleles that retain these regulatory regions . As we have shown , alternative transcripts may escape nonsense-mediated decay so 1 ) analyze the DNA sequence for nearby cryptic splice sites , especially those in frame to the natural/altered cryptic splice site , 2 ) check whether a nonsense mutation is in a predicted splicing enhancer sequence using available web tools , and 3 ) in the case of expected exon skip , analyze the exonic sequence for essential motifs and whether the exon length is divisible by 3 . Since shorter introns that precede expected affected exons may be retained ( instead of exon skip ) , intron length is also a factor to consider when generating mutants . Performing these steps near the start of a project can inform the nature and location of mutations that would most likely result in a loss-of-function mutant with a phenotype of interest .
All procedures using zebrafish were approved by the Carnegie Institution Animal Care and Use Committee ( Protocol# 139 ) or the Institutional Animal Care and Use Committee of the University of Maryland ( Permit Number: 0610009 ) . All lines were raised and crossed according to zebrafish husbandry guidelines [41] . Heterozygotes for each mutation were identified through a fin-clip based gDNA isolation ( REDExtract-N-Amp Tissue PCR kit; Sigma-Aldrich ) , PCR amplification of a 400–600 bp region around the mutation using designed primer sets ( MacVector , Primer 3 ) , and Sanger sequencing using a nested sequencing primer . ( Primer sets and conditions are in S3 Table . ) For the creb3l3asa18218 line , an NaOH-based DNA extraction method was used to extract gDNA from fin tissues . Genotyping primers were designed using dCAPS finder 2 . 0 with one mismatch ( http://helix . wustl . edu/dcaps/dcaps . html; [42] ) . The primer introduces EcoRV restriction sites in the mutant amplicons but not in the WT amplicons . The genomic location of each mutation is based on Ensembl genome assembly GRCz10 and calculated according to the guidelines of the Human Genome Variation Society . For each line with a mutation in an ESS , larvae were collected from incrosses of identified heterozygotes and 10–20 6-dpf larvae were pooled for generating RNA samples ( using above protocol ) . RNA samples served as template to generate cDNA ( iScript cDNA Synthesis Kit , Bio-Rad ) . cDNA samples were PCR amplified to provide amplicon sizes of 400–700 bp ) and the products were separated and sized using gel electrophoresis . For lines that showed evidence of a skipped exon , individual larvae were genotyped and treated similarly to above to correlate amplicon size with genotype . Adults that were found to carry the ESS mutation were incrossed and the progeny collected . Individual 6-dpf larvae underwent a Trizol-based RNA prep adapted from Macedo and Ferreira ( 2014 ) [43] to include an additional chloroform extraction . To genotype individual samples , residual gDNA in the unpurified RNA samples was PCR amplified and sent for Sanger sequencing . After genotypes were determined ( SnapGene Viewer to view peak trace files ) , RNA samples were DNAase I-treated and purified ( RNA Clean and Concentrator , Zymo Research ) , served as templates for cDNA synthesis ( iScript cDNA Synthesis Kit , Bio-Rad ) , and ultimately used in qPCR studies to analyze transcript levels . qPCR methods included SYBR Green-based methods ( Sigma-Aldrich , abca1a , creb3l3a , smyd1a ) and Taqman gene expression assays ( ThermoFisher Scientific; cd36 , slc27a2a , and abca1b ) . ef1α ( for smyd1a ) or 18s rRNA ( for all others ) levels were used as reference genes . Primer and assay information is shown in S2 Table . cDNA samples for individual larvae , along with “No RT” controls and “No transcript” controls were run on the CFX96 Touch Real-Time PCR Detection System ( Bio-Rad ) or on the 7500 Fast-Time PCR System ( AB Applied Biosystem ) . Three technical replicates were run for each sample and a minimum of three biological replicates were used for each genotype for each line . Data was analyzed through calculation of Delta Ct values ( 18s rRNA as internal control ) and either one-way analysis of variance ( ANOVA ) with the Tukey post hoc test or the Wilcoxon-Mann-Whitney test [44] . Transcript counting data for the pla2g12bsa659 mutant line was obtained from the Sanger Zebrafish Mutation Project [1] and processed as described [36] . Differential expression analysis was performed using DESeq2 ( 2010 ) [45] . The data are deposited at ENA under accession number ERP004581 ( samples ERS401972-ERS401991 ) . cDNA for individual larvae was generated as described above and sequencing was obtained using Sanger sequencing methods . Peak trace files were analyzed manually using SnapGene Viewer ( GSL Biotech , LLC ) or MacVector . To assist in determining the two alleles of interest ( wildtype and potential mutant ) for each line , Poly Peak Parser [46] and alignment of wildtype and mutant alleles in MacVector ( Align to Reference ) were used . The smyd1a allele containing a 7-bp deletion was generated using the CRISPR/Cas9 targeting method ( Cai and Du , in preparation ) . The target site ( 5’-GGACCTGAAGGAGCTCAAA-3’ ) was located in exon 3 of the smyd1a gene . Genotyping was carried out by using gDNA extracted from the caudal fin as a template for PCR followed by SacI digestion of the resulting amplicons . The 7-bp deletion abolished the SacI site , allowing resolution of bands by agarose gel electrophoresis . Homozygous smyd1a mutants were identified by PCR and SacI digestion , and confirmed by sequencing the PCR product . To compare the smyd1a mRNA sequences from wild type ( WT ) and mutant embryos , total RNA was isolated from a pool of 50 WT and homozygous smyd1a mutant embryos at 48 hpf . cDNAs were generated using the RevertAid First Strand cDNA Synthesis Kit ( ThermoFisher , K1621 ) . The full length smyd1a cDNA was amplified from the WT and mutant template using Phusion High-Fidelity DNA Polymerase ( NEB , M0530S ) . The amplicons were A-tailed using Taq DNA polymerase ( Promega , M8295 ) and subsequently cloned into pGEM-T easy ( Promega , A1360 ) . Length of all feature-tagged exons ( 1–10 ) in Ensembl genome assembly GRCz10 ( Release 90; August 2017 ) were divided by 3 and remainder ( 0 , 1 , or 2 ) noted . Number of exons analyzed were as follows: for exon 1 , 62895; 2 , 57137; 3 , 48568; 4 , 41336; 5 , 35312; 6 , 30144; 7 , 25617; 8 , 21969; 9 , 19057; 10 , 16738 . For each exon number , percentage of remainders 0 , 1 , and 2 were calculated . Chi-squared test for given probabilities was performed using the R Project for Statistical Computing v3 . 3 . 1 .
|
The recent rise of reverse genetic , gene targeting methods has allowed researchers to readily generate mutations in any gene of interest with relative ease . Should these mutations have the predicted effect on the mRNA and encoded protein , we would expect many more abnormal phenotypes than are typically being seen in reverse genetic screens . Here we set out to explore some of the reasons for this discrepancy by studying seven separate mutations in zebrafish . We present evidence that thorough cDNA sequence analysis is a key step in assessing the likelihood that a given mutation will produce hypomorphic or null alleles . This study reveals that mRNA processing in the mutant background often produces transcripts that escape nonsense-mediated decay , thereby potentially preserving gene function . By understanding the ways that cells avoid the deleterious consequences of mutations , researchers can better design reverse genetic strategies to increase the likelihood of gene disruption .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"fish",
"vertebrates",
"animals",
"animal",
"models",
"osteichthyes",
"mutation",
"developmental",
"biology",
"model",
"organisms",
"nonsense",
"mutation",
"experimental",
"organism",
"systems",
"sequence",
"motif",
"analysis",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"genome",
"complexity",
"genomics",
"artificial",
"gene",
"amplification",
"and",
"extension",
"bioinformatics",
"life",
"cycles",
"molecular",
"biology",
"zebrafish",
"point",
"mutation",
"eukaryota",
"polymerase",
"chain",
"reaction",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"larvae",
"introns",
"organisms"
] |
2017
|
mRNA processing in mutant zebrafish lines generated by chemical and CRISPR-mediated mutagenesis produces unexpected transcripts that escape nonsense-mediated decay
|
The bacterium Deinococcus radiodurans is one of the most radioresistant organisms known . It is able to reconstruct a functional genome from hundreds of radiation-induced chromosomal fragments . Our work aims to highlight the genes involved in recombination between 438 bp direct repeats separated by intervening sequences of various lengths ranging from 1 , 479 bp to 10 , 500 bp to restore a functional tetA gene in the presence or absence of radiation-induced DNA double strand breaks . The frequency of spontaneous deletion events between the chromosomal direct repeats were the same in recA+ and in ΔrecA , ΔrecF , and ΔrecO bacteria , whereas recombination between chromosomal and plasmid DNA was shown to be strictly dependent on the RecA and RecF proteins . The presence of mutations in one of the repeated sequence reduced , in a MutS-dependent manner , the frequency of the deletion events . The distance between the repeats did not influence the frequencies of deletion events in recA+ as well in ΔrecA bacteria . The absence of the UvrD protein stimulated the recombination between the direct repeats whereas the absence of the DdrB protein , previously shown to be involved in DNA double strand break repair through a single strand annealing ( SSA ) pathway , strongly reduces the frequency of RecA- ( and RecO- ) independent deletions events . The absence of the DdrB protein also increased the lethal sectoring of cells devoid of RecA or RecO protein . γ-irradiation of recA+ cells increased about 10-fold the frequencies of the deletion events , but at a lesser extend in cells devoid of the DdrB protein . Altogether , our results suggest a major role of single strand annealing in DNA repeat deletion events in bacteria devoid of the RecA protein , and also in recA+ bacteria exposed to ionizing radiation .
The extreme resistance of the bacterium D . radiodurans to DNA-fragmenting treatments , such as ionizing radiation or desiccation , is correlated with the ability to reconstruct a functional genome from hundreds of chromosomal fragments . The rapid reconstitution of an intact genome is thought to occur through an extended synthesis-dependent strand annealing process ( ESDSA ) followed by DNA recombination [1 , 2] . During ESDSA , chromosomal fragments with overlapping regions are used both as primers and templates for a massive synthesis of single-stranded DNA extensions . Newly synthesized complementary single stranded DNA extensions appear to anneal so that contiguous DNA fragments are joined together forming long linear intermediates . These intermediates require RecA-dependent homologous recombination to mature into reconstituted circular chromosomes representing DNA patchworks of numerous double-stranded DNA blocks synthesized before irradiation connected by DNA blocks synthesized after irradiation . We have recently shown that the Deinococcal RecF , RecO , RecR proteins , by their ability to load RecA onto its single-stranded DNA substrate , play a crucial role in DNA double strand break repair via ESDSA and recombinational repair pathways [3] . Mutant ΔuvrD bacteria showed a markedly decreased radioresistance , an increased latent period in the kinetics of DNA double strand break repair and a slow rate of fragment assembly correlated with a slow rate of DNA synthesis , suggesting that UvrD helicase might be involved in the processing of double stranded DNA ends and/or in the DNA synthesis step of ESDSA [3] . More recently , it was proposed that a single strand annealing ( SSA ) process participates in an early step of DNA double strand break repair by facilitating the accurate assembly of small fragments to generate suitable substrates for subsequent ESDSA-promoted genome reconstitution [4] . The DdrB protein was previously shown to exhibit in vitro properties akin to those of SSB protein [5] but also to promote annealing of single stranded DNA [6] . The DdrB protein , recruited early after irradiation into the nucleoid , was also shown to be involved in the slow DNA double strand break repair observed in cells devoid of the RecA protein , and thus to play a major role in RecA-independent DNA double strand break repair through SSA [4 , 6] . Rearrangements between repeated sequences are a major source of genome instability and can be deleterious to the organism . These rearrangements can result in deletion or duplication of genetic material flanked by direct repeats . In Escherichia coli , recombination between directly repeated sequences occurs via both RecA-independent and RecA-dependent mechanisms , depending on the size of the repeats and of the intervening sequences between the repeated sequences [7–9] . Insertion of a sizable DNA sequence in between the repeated sequences substantially increased the RecA dependence , suggesting that increasing the distance separating the homologous regions preferentially inhibits the RecA-independent recombination in E . coli [9 , 10] . In E . coli , RecA-independent rearrangements between short repeats , such as deletions , are stimulated by mutations that affect the DNA polymerase or other proteins involved in DNA replication [11–13] leading to the proposal that these events occur during DNA replication by a mechanism involving mispairing of the newly synthetized DNA strand with an alternative complementary template sequence located nearby [7 , 12] ( for review , see [14 , 15] ) . An alternate mechanism for RecA-independent deletion events involves DNA breakage , exonucleolytic erosion of the DNA ends and single strand annealing ( SSA ) of exposed complementary single stranded DNA [14] . A single strand annealing mechanism has also been proposed for RecA-independent deletions associated with a restart of collapsed replication forks [7 , 11 , 12 , 14 , 16] . Here , we measured the frequency of recombination between direct repeats separated by intervening sequences of various lengths restoring a functional tetA gene in the presence or absence of radiation-induced DNA double strand breaks in D . radiodurans . We also assessed the involvement of the RecA , RecO , RecF , UvrD and DdrB proteins in the deletion process . The role of these proteins in the progression of replication forks was also discussed .
To investigate the role of recombination proteins in the occurrence of repeat-mediated deletion events in D . radiodurans , we constructed a mutated tetA allele bearing an internal duplication and a spc cassette inserted between the duplicated regions . The engineered tetA allele was inserted into the dispensable amyE locus of chromosome 1 and provided a recombination substate in which the direct repeats ( 438 bp long ) were separated by a 1 , 479 bp spacer ( Fig 1A ) . Precise deletion of one of the direct repeats and the intervening sequence restores the wild type tetA allele . The presence of a functional tetA gene on one copy of chromosome 1 suffices to confer tetracycline resistance to the cells , although D . radiodurans bacteria contain 4 to 10 genome equivalents per cell . In contrast , when the deletion of a gene generates a loss-of-function mutant , all the copies of the gene must be eliminated to detect the mutant phenotype , and failure to obtain a homozygote provides a first indication that the gene might encode a function essential for cell viability [3 , 17] . The frequency of spontaneous deletion events in a population was estimated by measuring the frequency of [TetR] bacteria . As shown in Fig 1B , the median of the frequencies of [TetR] in the wild type bacteria was 6 . 5x10-4 and did not decrease in ΔrecA bacteria ( median value: 8x10-4 ) . Does the apparent RecA-independent high frequency of deletion events result from a functional redondancy of RecA activities in the cells ? We tested the involvement of the RadA protein , a RecA-related protein , and showed that wild type frequencies of deletions were found in ΔradA and in ΔrecA ΔradA bacteria ( Fig 1B ) , suggesting that the RadA protein was not required in RecA-independent recombination to compensate for the absence of the RecA protein . We also observed that the frequency of [TetR] bacteria was not reduced in cells devoid of RecF or RecO proteins required for loading RecA onto its single-stranded DNA substrate ( Fig 1B ) . Altogether , our results suggest an important contribution of RecA-independent mechanisms in the generation of deletions between repeated DNA sequences in D . radiodurans . To test whether the same 438 bp homologous fragments can undergo efficient RecA-dependent recombination , we constructed a plasmid-by-chromosome recombination assay in which the recombining tetA fragments were placed in a different configuration , one being located at the chromosomal amyE locus and the other on a resident plasmid ( Fig 2A ) . In this assay , the reconstitution of a functional tetA gene resulted from the integration of the plasmid into the chromosomal DNA as verified by PCR analysis of few [TetR] colonies ( S1 Fig ) . The integration event is well tolerated by the cell , since we used a low copy number plasmid p15002 , a derivative of plasmid pI8 maintained at 4 to 10 copies per cell in D . radiodurans [18] . As can be seen in Fig 2B , the frequency of [TetR] bacteria dropped from a median value of 2x10-5 in the wild type to less than 5 x 10−7 in cells devoid of the RecA or the RecF proteins , indicating that the 438 bp homologous fragments recombine through a classical RecA-promoted strand exchange mechanism . In contrast , loss of the RadA protein did not impair recombination efficiency . In E . coli , the RadA protein has been involved in processing of branched recombination intermediates . However single radA mutants have a modest effect on recombination and DNA survival while they show a strong synergistic effect in combination with mutations in the recG or the ruvAB Holliday junction proteins [19 , 20] . The presence of mutations in one of the 438 bp repeats ( Fig 3A ) reduced the frequency of the deletion events between the chromosomal repeats . A single mutation sufficed to significantly decrease the frequency of the deletion events as shown by median values that decreased by a factor of 4 . 8 in recA+ bacteria and 4 . 6 in ΔrecA bacteria , when compared to fully homologous repeats . The decrease was greater when 3 mutations were present in one of the repeats , yielding reduction factors of 10 . 0 and 19 . 0 in recA+ and ΔrecA bacteria , respectively ( Fig 3B ) . A plot of the frequency of [TetR] bacteria as a function of the number of differences shows a linear decrease in the deletion frequency with similar regression slopes in recA+ and ΔrecA bacteria ( S2 Fig ) . A similar analysis was performed in recA+ ΔmutS and ΔrecA ΔmutS bacteria devoid of the MutS protein , the key enzyme involved in mismatch recognition . The data ( Fig 3B and S2 Fig ) show that in this case the differences between the repeats did not significantly affect the deletion frequency . These results suggest that , as in homologous recombination intermediates , a heteroduplex DNA is formed during RecA-independent processes leading to the reconstitution of a functional tet gene and that an efficient mismatch repair aborts recombination between the DNA repeats in ΔrecA as well as in recA+ bacteria . It was previously shown that uvrD mutations stimulate RecA-dependent recombination [21–23] and enhance tandem repeat deletions in E . coli [23] . Here , we show that the absence of UvrD enhanced the efficiency of RecA-dependent recombination between chromosomal and plasmid DNA by a factor of 21 . 2 ( Fig 2B ) , suggesting that deinococcal UvrD protein possesses an anti-RecA activity as previously shown for the E . coli UvrD protein [24 , 25] . The absence of UvrD also enhanced the frequency of deletions between the chromosomal direct repeats by a factor of 7 . 1 ( Fig 1B ) . This increase may be due to the anti-RecA activity of UvrD protein that can possibly inhibit RecA-dependent recombination between the repeated sequences in a recA+ uvrD+ background . However , the absence of UvrD might also disturb DNA replication , and thus increase genome instability . A clue to understand how the absence of the UvrD protein might be involved , independently of its anti-RecA activity , in the stimulation of deletion events requires an analysis of its effects in a recombination-deficient background . Unfortunately , we were unable to obtain homozygotes for recA , recF or recO deletion in combination with a uvrD deletion , even after extensive purification steps ( S3 Fig ) , suggesting that uvrD deletion is colethal with a recA , recF or recO deletion . We propose that the UvrD protein , by displacing obstacles downstream of the replisome , plays an important role in the progression of replication forks ( see discussion ) . Previous in vitro and in vivo results suggest that the DdrB protein plays a major role in a single strand annealing process ( SSA ) that operates early in genome reconstitution after DNA damage [4 , 6] . Single strand annealing is the only activity of DdrB known besides binding to single strand DNA [5 , 6] . Thus , to analyse the involvement of SSA in generating deletions via a RecA-independent pathway , we decided to construct double mutants devoid of DdrB and RecA or RecO proteins . Homogenotization of ΔddrB ΔrecA and ΔddrB ΔrecO double mutants was difficult , requiring 7 steps of purification , suggesting growth inhibition of the mutated cells ( S3 Fig ) . Thus , we compared the growth rate and the plating efficiency of the double mutants with those of the single ΔddrB , ΔrecA and ΔrecO mutants and of the parental wild type strain . Wild type and ΔddrB bacteria exhibited a generation time of 105 min . The recombination deficient bacteria grew more slowly , as ΔrecA and ΔrecO bacteria during exponential growth showed a generation time of 285 min whereas ΔddrB ΔrecO and ΔddrB ΔrecA exhibited a generation time of 370 min . Moreover , during exponential growth phase , the single ΔrecA and ΔrecO mutants had a 15 fold decreased plating efficiency as compared with the wild type , whereas the ΔddrB ΔrecA and ΔddrB ΔrecO double mutants had a 35 fold decreased plating efficiency ( Fig 4 ) . These results suggest that the DdrB protein may be involved in management of blocked replication forks in the absence of the RecA or RecO proteins . Another striking result was the increased lethality of the recombination deficient mutants in late stationary phase . Indeed , after reaching a plateau after 6 hours of incubation for the wild type and ΔddrB bacteria , the number of CFU did not decrease during 70 additional hours of incubation . In contrast , the single ΔrecA and ΔrecO recombination deficient mutants and the double ΔddrB ΔrecA and ΔddrB ΔrecO mutants reached a plateau after 18 to 20 hours of incubation and the number of CFU decreased 2–3 orders of magnitude after 30 hours of incubation ( Fig 4 ) suggesting that DNA lesions are generated during prolonged stationary phase and require recombination functions to be repaired . We found that the absence of DdrB had a strong negative effect on the frequencies of deletions events between the chromosomal repeats generated via a RecA-independent pathway . Indeed , when the repeats are separated by 1 , 479 bp , the median values of the deletion frequencies in ΔddrB ΔrecA and ΔddrB ΔrecO bacteria decreased by a factor of 4 . 9 and 5 . 1 respectively , as compared to their ΔrecA and ΔrecO counterparts ( Fig 1B ) . These results indicate that 80% of the [TetR] bacteria generated in the absence of RecA or RecO proteins were formed in a DdrB-dependent manner , suggesting a major role of single strand annealing in RecA-independent recombination between the direct repeats . We can also hypothesize that the DdrB protein might be involved in the stabilization of DNA polymerase template switching intermediates . In E . coli , in which both RecA-dependent and RecA-independent mechanisms can contribute to recombination between direct repeats , deletion events become increasingly RecA-dependent as the distance between the repeated sequences increases [9] . To verify if this also applies in D . radiodurans , we modified the test deletion construct shown in Fig 1 by replacing the 1 , 479 bp spacer with sequences of increasing length to analyse the impact of the distance between the repeats on the incidence of deletions and their genetic control ( Fig 5A ) . We found that the increase of the distance between the repeats from 1 , 479 bp ( Fig 1A ) up to 10 , 500 bp ( Fig 5A ) had no effect on the deletion frequency in recA+ as well as in ΔrecA or ΔrecF hosts ( compare the [TetR] frequencies in Figs 1B , 5B , 5C and 5D ) . Likewise , the distance between the repeats had no effect on the stimulation of deletion events by the absence of the UvrD protein ( Figs 1B , 5B , 5C and 5D ) . In contrast , the involvement of DdrB in the deletion events became more apparent when the distance between the repeats increased . Indeed , while a DdrB deficiency had no effect on the frequency of deletions in a recA+ background when the spacer between the repeats was 1 , 479 bp long , it produced a 2 to 3-fold decrease in the deletion frequency when the spacer length increased ( Figs 1B , 5B , 5C and 5D ) . When the ddrB deletion was associated with a recA deletion , the reduction factors were found to be between 18- and 20-fold if the length of the intervening sequences was ≥ 3 , 500bp as compared to their single ΔrecA counterparts ( Fig 5B , 5C and 5D ) . A similar effect of a ddrB deletion was also observed in cells devoid of the RecO protein ( Fig 5B , 5C and 5D ) . These results suggest that , in the absence of the RecA-promoted homologous recombination , approximately 95% of the recombination events were dependent on the DdrB protein and may be related to an SSA pathway . We used our deletion assay to analyze the impact of the presence of repeated sequence on the stability of the genome in γ-irradiated cells during the process of genome reconstitution . We showed that the frequency of repeat-induced deletions restoring a functional tetA gene increased as a function of the dose of γ-irradiation used ( Fig 6A ) . Thus , we further exposed the cells to 5 kGy γ-irradiation , a dose producing hundreds of DNA double strand breaks [26] . The repeats in the tested cells were separated either by 1 , 479 bp ( “short” spacer ) , 3 , 500 bp , 6 , 500 bp , or by 10 , 500 bp ( “long” spacer ) sequences . The deletion analysis was performed only in a recA+ background , since the extreme radio-sensitivity of ΔrecA or ΔrecF bacteria ( cell survival was less than 10−5 after exposure to 5 kGy ) precluded the inclusion of these cells in the genetic assay . The deletions were induced by exposure to γ-irradiation independently of the length of the spacers ( Fig 6B ) . As shown in Fig 6 , in the wild type bacteria , the deletion frequency increased by 11 . 9 fold and 5 . 9 fold after irradiation when the repeats are separated by the”short” spacer and the “long” spacer , respectively ( compare left panels in Fig 6C and 6D ) . In cells devoid of the UvrD protein , the frequency of deletions moderately increased after irradiation ( induction factors of 1 . 45 and 1 . 7 for the “short” and “long” spacer , respectively ) , likely because these cells already have an elevated spontaneous level of recombination in the absence of irradiation ( compare right panels in Fig 6C and 6D ) . In contrast , cells devoid of the DdrB protein showed a marked reduction in the induced levels of deletion events ( compare middle panels in Fig 6C and 6D ) , suggesting that single strand annealing might play an important role in the generation of the [TetR] recombinants during genome reconstitution after irradiation .
Here , we determined the frequencies of spontaneous and radiation-induced recombination between chromosomal direct repeats and investigated the role of RecA and of other key recombination and repair proteins in the occurrence of these events in D . radiodurans . We found that recombination events restoring a functional tetA gene occurred at a very high frequency . In a recA+ background , the median frequency of [TetR] cells was equal to 6 . 5 x 10−4 ( Figs 1 , 5B , 5C and 5D ) , values more than 10-fold higher than those measured in a study that used a similar substrate ( chromosomal 358 bp direct repeats separated by an intervening sequence of 850 bp ) to measure recombination in Helicobacter pylori , a bacterium known for its high recombination proficiency [29] . Moreover , inactivation of recA resulted in a 10-fold decrease in recombination between the direct repeats in H . pylori [29] . In contrast , introduction of a ΔrecA , ΔrecF or ΔrecO mutation in the D . radiodurans tester strains did not change the elevated recombination frequencies between the repeats ( Figs 1 , 5B , 5C and 5D ) . In E . coli and in B . subtilis , the distance between the repeats plays a key role in determining the mechanisms involved in the recombination processes , the efficiency of RecA-independent recombination decreasing sharply when the distance between the repeats increases [9 , 10 , 30] . No such proximity effect was observed in D . radiodurans . Indeed , the frequency of appearance of the recombinants in our assay remained elevated in ΔrecA ( and ΔrecF or ΔrecO ) bacteria as in the parental rec+ bacteria when the spacer between the repeats increased from 1 , 479 up to 10 , 500 bp ( Fig 5 ) . Our results suggest that , in the absence of RecA ( or in the absence of “facilitator” proteins required for loading RecA onto its single-stranded DNA substrate ) , alternate pathways ensure recombination between repeated sequences in D . radiodurans . These RecA-independent pathways do not necessarily predominate in recA+ bacteria , although we observed a similar frequency of recombinants in recA+ and ΔrecA bacteria . Indeed , ΔrecA ( or ΔrecF or ΔrecO ) bacteria had a low plating efficiency with less than 10% of cells able to form colonies . In E . coli , mutations in components of the DNA Pol III holoenzyme result in elevated levels of tandem repeat rearrangements , supporting the idea that RecA-independent recombination occurs during the process of chromosome replication [14] . A replication slipped misalignment model [14] proposed that a pause in DNA synthesis and dissociation of the polymerase from its template allows the nascent strand to translocate to a new pairing position . Slipped misalignment is thought to occur on single-stranded DNA and thus more frequently during lagging-strand synthesis [14] . The availability of single-stranded DNA on the lagging-strand template , and thus , the length of the Okazaki fragments , might constitute parameters that govern the efficiency of the deletion events and might explain the strong dependence of the deletion frequencies on the proximity of the repeated sequences [14] . Our findings that the distance between the repeats in the 1–11 kb range has no influence on the deletion frequency raise the question as to whether these events were generated through a slipped misalignment mechanism , thus implying the presence of very large single stranded DNA regions on the lagging strand template in D . radiodurans . A response to this question awaits better knowledge of the replication machinery in D . radiodurans and a determination of the average size of Okazaki fragments in this bacterium . We found that the frequencies of the deletion events in ΔrecA ΔddrB ( or ΔrecO ΔddrB ) bacteria were reduced 5-fold and 18- to 20-fold as compared with those measured in the single ΔrecA ( or ΔrecO ) mutant counterparts ( Figs 1B and 5 ) in strains containing repeats separated by 1479 bp and 3 , 500 to 10 , 500 bp , respectively . These results strongly suggest that the DdrB protein strongly stimulates RecA-independent recombination in D . radiodurans . The DdrB protein was shown to bind single stranded DNA [5] and to mediate in vitro fast annealing of complementary oligomers [6] . In vivo , the single strand annealing activity of DdrB is supported by its involvement in plasmid establishment during natural transformation [4] . Thus , we propose that RecA-independent recombination between direct repeats occurs mainly through a DdrB-dependent single strand annealing ( SSA ) pathway . SSA was first proposed to explain circularization of linear duplex phage DNA containing terminal repetitions by annealing complementary terminal single overhangs [31] . The SSA model was postulated later to take place in eukaryotic cells [32 , 33] where it is facilitated by RPA and RAD52 in a RAD51-independent manner [34 , 35] . SSA involves an initial DNA double strand break in the sequence between the duplications followed by the action of a 5’ to 3’ exonuclease to expose single stranded regions in both repeats that are subsequently aligned and annealed by the RAD52-RPA-ssDNA ternary complex . Annealed intermediates are then processed by digestion of the displaced single stranded DNA , polymerase filling-in and ligation to generate the final recombination product ( For review , see [36] ) . Although DdrB does not share sequence similarity with the eukaryotic RAD52 protein , it might act as its functional equivalent [37 , 38] . The activity of the RAD52 protein is strongly stimulated by the presence of RPA [39 , 40] . In contrast , the single strand annealing activity of DdrB is not stimulated but rather inhibited by inclusion of the SSB protein in the in vitro annealing reaction [4 , 6] . The deinococcal SSB protein is an essential protein and the DdrB protein is unable , even when overexpressed , to replace SSB for cell viability [41] . SSB is crucial for all aspects of DNA metabolism [42] while DdrB seems to have a more specialized role in DNA repair and plasmid transformation by stimulating the SSA pathway . Both DdrB and SSB bind to 3’ single stranded tails of resecting ends [5] . The ends can be engaged in two alternative pathways: annealing to complementary ssDNA in the SSA pathway , or , depending on the formation of a RecA nucleofilament , invasion of homologous dsDNA to promote strand exchange in homologous recombination or to prime DNA synthesis in ESDSA . Polymerization of RecA on ssDNA requires the displacement of SSB or DdrB from ssDNA . The SSB protein can be efficiently displaced through the action of RecO and RecR proteins [43 , 44] . DdrB protein binds more tightly than SSB to ssDNA [5] and might be displaced with more difficulty from ssDNA . We propose that , in D . radiodurans , homologous recombination and SSA might also compete for substrate in making deletions between direct repeats . Reams et al . [45] proposed that , in Salmonella enterica , a single-strand annealing pathway might also be activated to generate duplication between tandem copies of the ribosomal RNA genes ( rrn ) when two single-stranded DNA ends are provided and neither strand is coated with inhibitory RecA protein . Under these conditions , the activation of single-strand annealing might compensate the loss of homologous recombination [45] . We found that ΔrecA ( or ΔrecO ) bacteria had a 15 fold decreased plating efficiency as compared with the wild type during exponential growth phase ( Fig 4 ) . This important lethal sectoring suggests that problems resulting in the arrest of replication fork ( see for review [46] ) occur at high frequency in D . radiodurans , and that RecA-mediated recombination plays a key role in the recovery of stalled replication forks in this bacterium . A further ( 2-fold ) decreased plating efficiency was observed when a DdrB deficiency was combined with a RecA ( or RecO ) deficiency ( Fig 4 ) . These results are consistent with a single strand annealing model but also with any model that envisions annealing of complementary DNA strands , for example misannealing of the direct repeats during recovery from replication fork collapse in cells devoid of the RecA protein . D . radiodurans also seems to be very sensitive to prolonged stationary phase , with a rapid loss of cell viability when proteins involved in homologous recombination were absent , suggesting that DNA double strand breaks were generated in “old” cells and could not be repaired in the absence of RecA or RecO proteins . Another important feature was the increased recombination frequency between repeated sequences measured in a D . radiodurans mutant devoid of the UvrD protein at a level almost equivalent to those measured after irradiation of wild type bacteria ( Fig 6 ) . It was previously shown that uvrD mutations enhance tandem repeat deletion in the E . coli chromosome [23 , 47] and stimulate RecA-dependent recombination [21 , 48 , 49] . In E . coli , mutations in uvrD induce the SOS response , a common phenotype in cells with replication defects [50] . The obstacles possibly encountered by replication forks during their progression are multiple , such as tightly bound proteins , nicks or DNA lesions . The Rep helicase acts by dislodging proteins in front of replication forks [51–53] and its absence results in a marked slowing down of replication progression [54] , suggesting increased fork arrest . Simultaneous inactivation of Rep and UvrD helicases is lethal in E . coli [23] suggesting that UvrD might partially substitute for the Rep protein in ensuring replication progression [55] . In favour of this hypothesis , it was recently shown that UvrD displaces the obstacles downstream of the replisome in vitro [52] and plays a major role to displace transcription complexes [56] . Moreover , it was proposed that UvrD acts at blocked replication forks by clearing RecA , facilitating replication fork reversal [57 , 58] , a hypothesis supported by the ability of UvrD to directly remove RecA nucleoprotein filaments in vitro [25] . In D . radiodurans , we previously showed that inactivation of uvrD results in a marked slowing down of replication progression in un-irradiated cells [3] . During post-irradiation recovery through ESDSA , the absence of UvrD results in a delayed kinetics of DNA double strand break repair that coincided with delayed and less extensive DNA synthesis than that observed in the wild type cells [3] . D . radiodurans bacteria are naturally devoid of the RecB and RecC proteins , and it was suggested that UvrD , in association with the RecJ exonuclease , might play an important role in the processing of DNA double strand ends required for priming of DNA synthesis , but also may act in the DNA synthesis elongation step of ESDSA and more generally may play an important role for the progression of replication forks [3] . It is important to notice that we were unable to obtain mutants devoid of the RecJ protein [3] , and recJ mutants constructed by Hua and his collaborators were shown to grow very slowly and to be thermosensitive [59] . In D . radiodurans , we were unable to delete the recA gene when bacteria were devoid of the UvrD protein , suggesting colethality of uvrD and recA deficiencies . These results are reminiscent of phenotypes observed in particular rad3 mutants of Saccharomyces cerevisiae . The RAD3 gene , a homolog of the human gene XPD , encodes a helicase which is a component of the NER apparatus as part of the transcription factor TFIIH . Interestingly rad3-101 and rad3-102 mutants accumulate DNA double strand breaks and are lethal when in combination with mutations in recombinational repair genes , strongly suggesting that Rad3 protein influences either the generation of DNA double strand breaks or their processing by homologous recombination [60] . We propose that the absence of UvrD in D . radiodurans may disturb the progression of the replication fork , and thus might , as RAD3 in S . cerevisiae , influence the generation of DNA double strand break , favouring recombination and also single strand annealing between DNA repeats . We used our assay to analyze the impact of the presence of repeated sequence on the stability of the genome in γ-irradiated cells . We found that exposure to a dose of 5 kGy γ-irradiation increased the recombination level about 10-fold in the wild type but to a lesser extent in cells devoid of the DdrB protein , suggesting that SSA might play an important role in recombination between the duplicated sequences during the process of genome reconstitution . In D . radiodurans , interplasmidic recombination between homologous regions was previously shown to be induced by exposure to γ-radiation [61] . Moreover , when two TetS alleles were inserted on the same chromosome into two randomly distant sites , 2% of TetR bacteria were found among the surviving cells exposed to 17 . 5 kGy , whereas TetR isolates were only very rarely found without irradiation [62] . Interestingly , when two slightly different E . coli plasmids were inserted in the D . radiodurans genome generating adjacent duplication insertions , circular derivatives of the tandemly integrated plasmids were formed in the first 1 . 5 h postirradiation before the onset of recA-dependent repair in cells exposed to 17 . 5 kGy γ-irradiation . These circular derivatives had structures consistent with the hypothesis that DNA repair occurred immediately postirradiation by a recA-independent single strand annealing process [63] . These authors proposed that SSA may be a preparatory step for further DNA repair in wild-type D . radiodurans , a hypothesis in accordance with our recent results , suggesting that DdrB-dependent single-strand annealing might facilitate the assembly of the myriad of small fragments generated by extreme radiation exposure to generate suitable substrates for subsequent ESDSA-promoted genome reconstitution [4] . Genome reassembly in irradiated D . radiodurans cells was considered for a long time as an error-free process since no genome rearrangements were detected after post-irradiation DNA repair . Gross chromosomal rearrangements were detected for the first time in recA+ D . radiodurans cells exposed to extremely high γ-doses ( 25 kGy ) and in recA mutant cells that survived 5 kGy γ-radiation [64] . The recA mutants were also shown to be prone to spontaneous DNA rearrangements during normal exponential growth [64] . These authors presumed that SSA , by pairing ectopic repetitive sequences , may be the main source of these chromosomal rearrangements , a hypothesis reinforced by our results suggesting an important role of SSA in recombination between repeated sequences ( this work ) , in DNA double strand break repair in cells devoid of the RecA protein [4 , 6] , and in early reassembly of small DNA fragments when cells were exposed to high γ-doses [4] . Altogether , these results suggest that SSA plays a major role in RecA-independent recombination between repeated sequences in the radioresistant D . radiodurans bacterium . In un-irradiated wild type bacteria , the deletion events might result , as proposed by Susan Lovett in E . coli , from RecA-dependent intermolecular unequal crossing over or intramolecular recombination between the overlapping 5’ and 3’ regions of the tetA gene , and from RecA-independent processes such as replication slippage or template switching [10] or single strand annealing [10 , 14] . Difficulties in replication can lead to breakage of the fork when replication forks are halted by obstacles or DNA damage in virtually every cell and every cell generation [65 , 66] . If this occurs in the context of repeated DNA sequences , single-stranded DNA substrates might be generated by resection of the DNA ends , and genetic rearrangements can result through strand-invasion of the broken chromosome with its sister or through SSA at the repeats . Replication of damaged DNA templates can further elevate the probability of fork breakage [67 , 68] . Moreover , when D . radiodurans cells were exposed to a dose of 5 kGy γ-irradiation , generating hundreds DNA double strand breaks , DdrB-dependent SSA and RecA-dependent ESDSA processes involved in DNA double strand break repair increased the opportunities to generate deletion events when DNA repeats are present in the DNA fragments .
D . radiodurans strains were grown at 30°C , 150 rpm in TGY2X ( 1% tryptone , 0 . 2% dextrose , 0 . 6% yeast extract ) or plated on TGY1X containing 1 . 5% agar . E . coli strains were grown at 37°C , 150 rpm in Lysogeny Broth ( LB ) . When necessary , media were supplemented with the appropriate antibiotics used at the following final concentrations: kanamycin , 6 μg/mL; chloramphenicol , 3 . 5 μg/mL; hygromycin , 50 μg/mL; spectinomycin , 75 μg/mL; tetracycline 2 . 5 μg/mL for D . radiodurans and kanamycin , 25 μg/mL or spectinomycin 40 μg/mL for E . coli . The bacterial strains and plasmids used in this study are listed in Table 1 . The E . coli strains used were DH5α as the general cloning host , and SCS110 , a dam dcm mutant strain , to propagate plasmids prior introduction into D . radiodurans via transformation [69] . Transformation of D . radiodurans with genomic DNA , PCR products , or plasmid DNA was performed as described [70] . All D . radiodurans strains were derivatives of strain R1 ATCC 13939 . The genetic structure and the purity of the mutant strains were checked by PCR . Oligonucleotides used for strain constructions and diagnostic PCR will be provided on request . Cells were plated on TGY agar and incubated at 30°C for 3 days , or 5 days for ΔrecA , ΔrecO and ΔrecF mutant bacteria and 7 days for double mutant ΔddrB ΔrecA and ΔddrB ΔrecO bacteria . Three to six colonies per strain were inoculated in 3 mL of TGY2X and incubated at 30°C , 150 rpm . Appropriate dilutions of the bacterial cultures , grown to OD650nm = 1 . 5 were plated on TGY and TGY + tetracycline 2 . 5 μg/mL . Colonies were counted after 4 to 7 days of incubation at 30°C . The experiments were repeated at least three times , using , when possible , strain isolates obtained independently during the strain constructions . To measure recombination between repeated sequences after γ-irradiation , bacterial strains were treated as previously described upstream , except that bacterial cultures were grown to an OD650nm = 0 . 5 before being concentrated by centrifugation in TGY2X to an OD650nm = 10 and irradiated on ice at a dose of 5 kGy with a 60Co irradiation system ( LABRA , CEA , Saclay ) at a dose rate of 100 Gy/min . Following irradiation , samples of 100 μL were inoculated in 4 . 9 mL of TGY2X and incubated at 30°C , 150 rpm . After 20 hours of post-irradiation incubation , appropriate dilutions of bacterial culture were plated on TGY and TGY + tetracycline 2 . 5 μg/mL . Colonies were counted after 4-7days of incubation at 30°C . Unirradiated controls were treated as irradiated cells , except that they were maintained on ice without irradiation during the period when the irradiated cells were exposed to γ-rays . Cells containing plasmid p15002 were plated on TGY agar + spectinomycin ( 75 μg/mL ) and incubated at 30°C during 3 days ( or 5 days for ΔrecA and ΔrecF bacteria ) . Three colonies per strain were inoculated in 3 mL of TGY2X + spectinomycin ( 75 μg/mL ) and incubated at 30°C , 150 rpm . Appropriate dilutions of the bacterial cultures grown to OD650nm = 1 . 5 were plated on TGY and TGY+ tetracycline 2 . 5 μg/mL . Colonies were counted after 4 to 7 days of incubation at 30°C . Strains were streaked on TGY plates supplemented with the appropriate antibiotics . Independent colonies were inoculated in 3 mL TGY2X supplemented with the appropriate antibiotics ( only kanamycin for double mutants ) and grown at 30°C to an OD650nm = 1 . 5 . Cultures were then diluted 200 to 5 , 000 fold and grown overnight at 30°C to an OD650nm = 0 . 1 ( time 0 of the growth curves ) . Then , the OD650nm was measured and appropriate dilutions were plated on TGY plates at different times during 80 h of incubation at 30°C with agitation ( 150 rpm ) . Colonies were counted after 3 ( WT or ΔddrB bacteria ) , 5 ( ΔrecA or ΔrecO bacteria ) or 7 days ( ΔddrB ΔrecA or ΔddrB ΔrecO bacteria ) of incubation at 30°C . In Figs 1 and 3 and 5 , in order to establish the statistical differences between the [TetR] frequencies measured in mutant and WT strains , non parametric Dunn’s multiple comparison test [75] were used , taking into account the p-value correction and performed with the GraphPad Prism6 software . All the comparisons were bi-sided . Linear regressions and the slope significances observed in S2 Fig were estimated using the GraphPad Prism6 software . In Fig 6 , statistically significant differences between the irradiated and the non-irradiated conditions were calculated by non-parametric Mann-Withney tests performed in GraphPad Prism 6 software .
|
Deinococcus radiodurans is known for its exceptional ability to tolerate exposure to DNA damaging agents and , in particular , to very high doses of ionizing radiation . This exceptional radioresistance results from many features including efficient DNA double strand break repair . Here , we examine genome stability in D . radiodurans before and after exposure to ionizing radiation . Rearrangements between repeated sequences are a major source of genome instability and can be deleterious to the organism . Thus , we measured the frequency of recombination between direct repeats separated by intervening sequences of various lengths in the presence or absence of radiation-induced DNA double strand breaks . Strikingly , we showed that the frequency of deletions was as high in strains devoid of the RecA , RecF or RecO proteins as in wild type bacteria , suggesting a very efficient RecA-independent process able to generate genome rearrangements . Our results suggest that single strand annealing may play a major role in genome instability in the absence of homologous recombination .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Single Strand Annealing Plays a Major Role in RecA-Independent Recombination between Repeated Sequences in the Radioresistant Deinococcus radiodurans Bacterium
|
The four dengue viruses , the agents of dengue fever and dengue hemorrhagic fever in humans , are transmitted predominantly by the mosquito Aedes aegypti . The abundance and the transmission potential of Ae . aegypti are influenced by temperature and precipitation . While there is strong biological evidence for these effects , empirical studies of the relationship between climate and dengue incidence in human populations are potentially confounded by seasonal covariation and spatial heterogeneity . Using 20 years of data and a statistical approach to control for seasonality , we show a positive and statistically significant association between monthly changes in temperature and precipitation and monthly changes in dengue transmission in Puerto Rico . We also found that the strength of this association varies spatially , that this variation is associated with differences in local climate , and that this relationship is consistent with laboratory studies of the impacts of these factors on vector survival and viral replication . These results suggest the importance of temperature and precipitation in the transmission of dengue viruses and suggest a reason for their spatial heterogeneity . Thus , while dengue transmission may have a general system , its manifestation on a local scale may differ from global expectations .
The dengue viruses are the most widely distributed and damaging arthropod-borne viruses ( arboviruses ) affecting humans . The viruses and their predominant mosquito vector , Aedes aegypti , are endemic to most of the tropical and subtropical regions of the world , where they cause seasonal epidemics of varying size . The seasonal nature of transmission may reflect the influence of climate on the transmission cycle . Increases in temperature and precipitation can lead to increased Ae . aegypti abundance by increasing their development rate , decreasing the length of reproductive cycles , stimulating egg-hatching , and providing sites for egg deposition [1] , [2] , [3] , [4] . Higher temperature further abets transmission by shortening the incubation period of the virus in the mosquito [5] . Theoretical models of dengue transmission dynamics based on mosquito biology support the importance of temperature and precipitation in determining transmission patterns [6] , [7] , but empirical evidence has been lacking . On global scales , several studies have highlighted common climate characteristics of areas where transmission occurs [8] , [9] , [10] . Meanwhile , longitudinal studies of empirical data have consistently shown that temperature and precipitation correlate with dengue transmission but have not demonstrated consistency with respect to their roles [11] , [12] , [13] , [14] , [15] , [16] , [17] , [18] , [19] , [20] , [21] , [22] . For example , cumulative monthly rainfall and mean temperature correlated positively with increased dengue transmission on the Andaman Sea side of Southern Thailand [17] . On the Gulf of Thailand side , however , it was the number of rainy days ( regardless of quantity ) and minimum temperature that associated positively with incidence . Another study , farther north , in Sukhothai , Thailand , found that temperature had a negative effect on dengue transmission [16] . This finding only makes biological sense at the upper temperature limits of Ae . aegypti survival , an uncommon condition during the study period . Other biologically suspect findings include a model for Selangor , Malaysia where lagged precipitation was a significant predictor of early wet season dengue , but did not associate with a significant change in Ae . aegypti abundance [12] , the theoretical mechanism for precipitation increasing transmission . Though the vector and fundamental transmission cycle are similar in all endemic areas , the described relationships between transmission and weather are highly variable and , in some cases , make little biological sense . Some of the different findings may be attributed to underlying climate heterogeneity; local mosquito populations may be limited by different aspects of the environment depending on the conditions that they experience . However , local differences may also be attributed to over-fitting and incomplete statistical control for autocorrelation and collinearity . Autocorrelation and collinearity pose significant challenges as they have important implications for regression models . Autocorrelation arises as a natural feature of infectious disease systems as the number of new infections relates closely to the number of recent infections . In longitudinal regression analysis this generally results in correlated residuals . Collinearity presents its own problem . Weather variables and dengue incidence are strongly seasonal , but seasonality alone does not imply a meaningful association , particularly when the effects may occur over lags of weeks or even months . The scale of seasonal variability in this system is so high that any lagged seasonal weather variable can account for a large proportion of the variation in dengue incidence . Here we analyzed the association of temperature and precipitation with dengue transmission in each of 77 municipalities of Puerto Rico over a 20 year period using adaptive natural cubic splines to adjust for seasonal confounding [23] . The only excluded municipality was Culebra , a separate island where transmission is relatively sporadic due to a very small resident population . Figure S1 shows monthly temperature , precipitation , and dengue incidence for three example municipalities . We used a hierarchical statistical model to examine local associations over time and spatial heterogeneity in the estimated local associations [24] . At the first stage , within each municipality , we estimated the local short-term association between monthly variation in weather variables and monthly variation in dengue incidence while controlling for the smooth seasonal pattern of each covariate and reducing autocorrelation in the residuals . More specifically , we fitted municipality-specific Poisson regression models with monthly dengue incidence regressed on monthly average temperature or precipitation with a population offset and a natural cubic spline function of time . Because there are inherent delays between weather , its impact on mosquito populations , and their subsequent impact on transmission patterns , we used distributed lag models to assess effects of weather on dengue transmission up to 6 months later [25] . In the second stage , we estimated global association by averaging the short-term associations across municipalities and identified local climate characteristics that modify the local short-term associations . This stage of analysis allows us to characterize the spatial heterogeneity of the relationship between weather and dengue transmission .
Clinically suspected dengue infections are reported to the surveillance system maintained by the Puerto Rico Department of Health and the Centers for Disease Control and Prevention . Though suspected and laboratory-confirmed cases generally correlate highly , we use suspected cases because they are more sensitive given that approximately 60% of reported cases have inadequate samples for a definitive diagnosis . Here we analyze monthly totals by municipality from July 1986 through December 2006 . Municipal population data was obtained from the 1980 [26] , 1990 [27] , and 2000 [28] United States censuses . Estimates for median household income and the percentage of individuals living below the poverty line for each municipality are from the 2000 census [29] . Monthly mean maximum temperature , mean minimum temperature , and cumulative precipitation are simulated on 1 km2 grids from cross-validated spatial models of weather station data [30] . Monthly average temperature is the mean of the monthly maximum and minimum . Monthly values of each weather variable are aggregated at the municipal level as an average of all internal pixels weighted by pixel population size [31] . This weighting adjusts the average to better reflect climate in the areas where people live and dengue transmission occurs . The number of reported dengue cases , y , at time t for each municipality is modeled as , The population size , Nt , is assumed to be a linear function of time parameterized by the 1980 , 1990 , and 2000 censuses [26] , [27] , [28] . The longitudinal covariates , x1 , … , xP , are entered at covariate-specific distributed lags , lp [25] . The natural cubic spline smoothing function of time , s ( t , λ ) , is assigned λ of 2 degrees of freedom per year to fit a curve with a smooth seasonal pattern . Parameter estimates , βp from each local regression model are compared using two-level normal independent sampling estimation [24] . First-stage location-specific ( j ) parameter estimates are distributed , The variance , σj2 , can be estimated from first-level models . However , because the covariates are also modeled , and thus have estimated intrinsic variance , we repeated the regression model for each of 1 , 000 weather model conditional simulations and use the distribution of these to estimate σj2 . Finally we added effect modifiers z1 , … , zQ , to estimate the average effects ( α0 ) and effect modification ( αq ) in a Bayesian model , All analyses were performed in R ( R Foundation for Statistical Computing , Vienna , Austria , 2007 ) .
In the distributed lag model including temperature at lags of 0 , 1 , and 2 months and precipitation at lags of 1 and 2 months , monthly variation in temperature was positively associated with monthly variation in dengue incidence in most municipalities ( Figure 1 ) . The global association ( average short-term association across all municipalities ) was positive and statistically significant at all three lags . Short-term associations were significant for monthly maximum and minimum temperature but weaker than those observed for average monthly temperature . Monthly variation in cumulative precipitation was significantly associated with monthly variation in dengue incidence in some , but not all , municipalities at lags of 1 and 2 months ( Figure 1 ) . Globally , this association was significantly positive only after accounting for local effect modification by long-term climate . These findings were robust to the addition of further temperature and precipitation lags . Long-term climate variables modified the local short-term association between monthly weather and dengue incidence for all variables at all lags . Long-term mean temperature significantly modified the short-term effect of monthly temperature and long-term mean precipitation significantly modified the effect of short-term precipitation . In municipalities with higher long-term temperature or precipitation , the short-term association between temperature or precipitation , respectively , and dengue incidence was weaker . For instance , the model predicts that an area with long-term mean temperature of approximately 30°C will exhibit no significant association between monthly variations in temperature and dengue incidence . This threshold was consistent across lags ( 30 . 6°C , 30 . 2°C , and 29 . 9°C for 0 , 1 , and 2 month lags , respectively ) and is compatible with laboratory studies which indicate that dengue transmission is optimized at temperatures above 30°C [5] . The effect of monthly precipitation is likewise minimized where mean annual precipitation is high , approximately 1 , 800 mm in Puerto Rico ( 1 , 820 mm and 1 , 760 mm , for 1 and 2 month lags , respectively ) . This condition is present in some areas of Puerto Rico and explains the lack of significant association between precipitation and dengue incidence in some municipalities . These thresholds must be considered as specific to Puerto Rico because they are contingent upon regional climate . High annual precipitation , for instance , may indicate consistently high precipitation or alternatively , a brief period of intense precipitation . In the latter case , even an area with high annual precipitation may exhibit a strong association with dengue transmission on the monthly scale . In addition to long-term mean climate measures , we analyzed effect modification associated with socio-economic factors including population density , median household income , and the percentage of families living below the poverty line . In municipalities with a higher poverty index the short-term association between weather variables and dengue incidence was stronger , but this effect was not consistent across lags . For a small island , Puerto Rico contains remarkable climate diversity: the northeastern coastal area is warm and wet; the central mountains , cooler and wetter; and the southwestern coastal region , hot and dry ( Figure 2 , Figure S1 ) . The effects observed here demonstrate how these differences influence the association between weather and dengue transmission . Figure 3 shows the cumulative effects of an increase in monthly temperature and precipitation on dengue incidence in the same month and in the following 1 to 2 months in each municipality . The cumulative effects were obtained by summing the short-term associations estimated at each lag . As expected , the cumulative effect of temperature on dengue incidence is highest in the cooler mountainous areas . Likewise , the role that precipitation plays is greatest in the dry southwestern coastal region . Regional patterns in transmission may result from both from shared characteristics and from the movement of infected humans or vectors among proximal municipalities . The spatial patterns observed in Figure 3 demonstrate regional behaviour due to climate . Though there may be additional effects of proximity , those described here are robust as they are derived based solely on local characteristics and only later compared at the global scale . Analysis of further spatial correlation due to movement requires more detailed mechanistic models beyond the scope of the current analysis .
The associations between temperature , precipitation , and dengue transmission reported here are strong and consistent through time . Moreover , these associations depend on local characteristics and have a biological interpretation . Together these associations suggest an important relationship . It is critical , however , to consider the extent of the role which temperature and precipitation may play in increasing dengue incidence . The spline smooth in the current analysis reduces the extensive inter-annual variation in incidence observed in endemic areas like Puerto Rico such that the analysis effectively isolates association on finer , monthly , temporal scales . Thus , while we have reported a significant association between climate and dengue incidence , it is on a month-to-month time scale and does not show that warmer years ( for example ) consistently exhibit higher overall incidence . Though temperature and precipitation may also influence the magnitude of yearly transmission , this analysis does not demonstrate that . Studies on the relationship between multi-year climate variation and dengue incidence are inconsistent and at most account for only part of inter-annual variation in dengue transmission [11] , [21] , [32] , [33] , [34] , [35] , [36] ( Puerto Rico manuscript in preparation; M . A . J . , D . A . T . Cummings , & G . E . G . ) . Cogent alternative hypotheses suggest the importance of intrinsic factors related to the interactions of the four serotypes of dengue virus with human populations [37] , [38] , [39] , [40] . The spline itself , being seasonal by definition , likely contains more variation that in reality is attributable to weather . Removing this variation from the analysis is critical to differentiating the effects of the covariates which also exhibit smooth seasonal trends . This makes the associations evident but may underestimate the magnitude of the true effects of temperature and precipitation . As described above in terms of inter-annual variation , the spline makes forecasting even on the monthly scale problematic . Although this limits the public health utility of our findings , the empirical demonstration of these associations and their dependence on the underlying climate are important for the understanding of transmission dynamics and the potential effects of changing climate . Our findings suggest that in areas where temperature and precipitation are already high , increases in either will have little effect on transmission . Spatial heterogeneity in transmission is a common feature of vector-borne pathogens and many other infectious diseases . Because of this , transmission may be more completely described by research focused on a local scale . However , such studies lack generalizability as the factors limiting transmission may not be universal . Large scale hierarchical studies containing spatial heterogeneity and reasonable consideration of confounding have the potential to reveal universal underlying patterns while simultaneously describing unique local conditions . The two-stage analysis employed here to estimate average global effects and determinants of local variation is an approach that could be applied to a number of other environmentally-mediated diseases . Previous studies of the empirical relationship between weather and local dengue transmission highlighted local differences and identified global climate characteristics of areas where transmission has been reported . Here , for the first time , we link the local , temporal relationship between weather and dengue transmission to underlying climate characteristics to account for heterogeneity in local transmission patterns .
|
Dengue viruses are a major health problem throughout the tropical and subtropical regions of the world . Because they are transmitted by mosquitoes that are sensitive to changes in rainfall and temperature , transmission intensity may be regulated by weather and climate . Laboratory studies have shown this to be biologically plausible , but studies of transmission in real-life situations have been inconclusive . Here we demonstrate that increased temperature and rainfall are associated with increased dengue transmission in subsequent months across Puerto Rico . We also show that differences in local climate within Puerto Rico can explain local differences observed in the relationship between weather and dengue transmission . This finding is important because it suggests that the determinants of transmission occur on a local level such that although dengue viruses have a basically universal transmission cycle , changes in temperature or rainfall may have diverse local effects .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"public",
"health",
"and",
"epidemiology/epidemiology",
"infectious",
"diseases/viral",
"infections",
"public",
"health",
"and",
"epidemiology/global",
"health",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"ecology/spatial",
"and",
"landscape",
"ecology",
"mathematics/statistics",
"ecology/ecosystem",
"ecology"
] |
2009
|
Local and Global Effects of Climate on Dengue Transmission in Puerto Rico
|
Glossina ( G . ) spp . ( Diptera: Glossinidae ) , known as tsetse flies , are vectors of African trypanosomes that cause sleeping sickness in humans and nagana in domestic livestock . Knowledge on tsetse distribution and accurate species identification help identify potential vector intervention sites . Morphological species identification of tsetse is challenging and sometimes not accurate . The matrix-assisted laser desorption/ionisation time of flight mass spectrometry ( MALDI TOF MS ) technique , already standardised for microbial identification , could become a standard method for tsetse fly diagnostics . Therefore , a unique spectra reference database was created for five lab-reared species of riverine- , savannah- and forest- type tsetse flies and incorporated with the commercial Biotyper 3 . 0 database . The standard formic acid/acetonitrile extraction of male and female whole insects and their body parts ( head , thorax , abdomen , wings and legs ) was used to obtain the flies' proteins . The computed composite correlation index and cluster analysis revealed the suitability of any tsetse body part for a rapid taxonomical identification . Phyloproteomic analysis revealed that the peak patterns of G . brevipalpis differed greatly from the other tsetse . This outcome was comparable to previous theories that they might be considered as a sister group to other tsetse spp . Freshly extracted samples were found to be matched at the species level . However , sex differentiation proved to be less reliable . Similarly processed samples of the common house fly Musca domestica ( Diptera: Muscidae; strain: Lei ) did not yield any match with the tsetse reference database . The inclusion of additional strains of morphologically defined wild caught flies of known origin and the availability of large-scale mass spectrometry data could facilitate rapid tsetse species identification in the future .
The trypanosomiasis infection risk of a particular area is determined by several factors , including tsetse species abundance and the sex distribution of a fly population [1] . While the sex is easily distinguishable with the bare eye , species identification can be challenging because there are 32 recognised tsetse species and subspecies [2] . Differentiation relies on morphological differences in colour , size and on minimal male genitalia variations [3] . Recent genome-based analyses revealed the subspecies status of seemingly uniform riverine G . palpalis palpalis individuals in Equatorial Guinea [4] . Accordingly , current tsetse specification based on morphology may not be the only way to rapidly determine the species status of Glossina spp . The matrix assisted laser desorption/ionisation time of flight mass spectrometry ( MALDI-TOF MS ) is an established method of identification for microorganisms [5] , [6] , [7] , [8] , [9] , [10] , [11] . The MALDI-based identification of microorganisms requires only a small portion of a microbial colony and a drop of matrix solution [12] , [13] , [14] . The intact microbial cells are mixed with matrix solution ( UV observing substances like alpha-Cyano-4-hydroxycinnamic acid , 2 , 5-dihydroxybenzoic acid ) , dried and subjected to laser induced soft ionization . The ions are then accelerated into a vacuum tube using a high electric field and the Time of Flight ( ToF ) to reach the detector is recorded . The velocity of an ion is inversely proportional to its mass , thus smaller ions travel faster than heavier ones and ions with the same charge travel together . The ions hitting the detector and their time of flight are visualized as spectra . The protein composition of each organism is unique , so a species-specific MALDI signature or spectrum is expected . The species identification does not require protein sequence data; instead the acquired spectra are matched with reference spectra database using a pattern- matching algorithm [9] , [11] . The technique proved to be time and cost effective , as reliable as genome-based identification methods [6] , [9] . Recently , MALDI-based species identification has been demonstrated for higher organisms as micro-algae , Prototheca [15] , [16] , the plant parasitic nematode Anguina tritici [17] , Drosophila [18] , [19] , ticks [20] biting midges ( Culicoides spp . ) [21] , [22] , [23] and fish [24] . In addition MALDI has also been utilised for differentiation of various eukaryotic cell lines [25] , immune cells [26] , [27] and for species level classification of ancient mammalian samples [28] . Several commercial software packages designed for microbial species identification are available and include , MALDI Biotyper ( Bruker Daltonics ) , the Axima ( Shimadzu ) -SARAMIS ( AnagnosTec ) systems ( now called VITEK MS ) ( BioMérieux ) , Andromas ( Andromas SAS ) systems and MicrobeLynx ( Waters ) [7] , [8] , [29] . As far as our knowledge is concerned , reference spectra data for insects or tsetse in particular have not been included in any of these software packages . We chose the MALDI Biotyper system for creating a tsetse-specific spectra database . This system calculates the log score value , or similarity score , by considering the matching proportion of the test spectra with the database reference spectra . It also considers the consistency of peak intensities among sample and reference spectra . The objective of this study was to investigate whether simple formic acid/acetonitrile extracts of five well known laboratory-reared tsetse breeds exhibit specific and reproducible peak patterns and if they prove to be valid for species level identification . Usually , field-collected tsetse are stored in ethanol and often parts of the insects are removed for diagnostics . Therefore , another goal was to investigate if any of the body parts ( head , thorax , abdomen , legs , wings and whole insects ) are useful for species prediction .
To establish a tsetse database , we utilised five well-established laboratory breeds listed in table 1 . They represent tsetse from three different habitats that are relevant for the transmission of trypanosomes that affect humans or animals [2] . Tsetse puparia were maintained at 26°C with a relative humidity of 75% . Two to 4 days after hatching they were sacrificed as tenerals at −18°C and then stored in ethanol ( 70% ) . A total of three insects each were obtained for the analysis of male and female entire individuals ( table 1 ) . Additionally , three males and females of each species were dissected representing the peak patterns of the body parts abdomen , head , legs , thorax and wings . The protein extraction was carried out as described in Murugaiyan et al . [16] . In brief , triplicates of each specimen ( whole insect , head , thorax , abdomen , wings and legs ) were washed with ethanol , air dried and mixed with equal volumes of 70% formic acid and 100% acetonitrile . The samples were then sonicated for 1 min on ice and the supernatants were collected for further analysis . One µl of each sample extract was spotted on to the MALDI target plate ( MSP 96 target polished steel ( MicroScout Target ) plate Bruker Daltonics , Bremen , Germany ) , dried and overlaid with 1 . 0 µl of saturated α-cyano-4-hydroxycinnamic acid matrix solution . The MALDI measurements were carried out using MALDI Microflex LT ( Bruker Daltonics , Bremen , Germany ) on a broad range of 2000–20000 m/z ( mass to charge ratio ) , following an external calibration with the bacterial test standard as recommended by the manufacturer . Each extract was spotted three times and each spot on the target plate was measured three times for acquiring 27 spectra per specimen . The spectra were acquired using the automated option ( AutoXecute acquisition mode ) in Flex control 3 . 0 software ( Bruker Daltonics , Leipzig , Germany ) . ( Box 1 ) In order to demonstrate the protein composition in each extract , Glossina ( G . ) palpalis gambiensis were chosen for an SDS-PAGE analysis [30] . In brief , the extracts of the whole insects and it's the body parts were precipitated in five volumes of ice-cold 100% acetone . The pellets were reconstituted with 10 µl of sample loading buffer , heated at 60°C for 5 minutes and separated using 4% stacking and 12% separating gel . The protein visualisation was carried out using Coomassie Blue staining [31] . Following the visual inspection using Flex analysis 3 . 0 software ( Bruker Daltonics , Bremen , Germany ) , the spectra were then loaded in Biotyper 3 . 0 ( Bruker Daltonics , Bremen , Germany ) software . The spectra were subjected to baseline subtraction ( multipolygonal; signal to noise ratio 3 ) and smoothing ( Savitzky Golay algorithm , frame size 25 Da ) . The composite correlation index [32] , a mathematical algorithm used to assess the spectra variations within and between each set of the measurements . The Composite Correlation Index ( CCI ) was computed using the standard settings of mass range 3000–12000 Da , resolution 4 , four intervals and autocorrelation off . The reference spectra were then created using the standard method version 1 . 2 settings of the software ( mass error of each single spectra: 2000 , desired mass error of main spectra: 200 , peak frequency: 25% and desired peak number: 70 ) . The cluster analysis ( main spectra dendrogram ) was calculated with “correlation” as distance measure and linkage at “complete” to evaluate the suitability of the MALDI-based differentiation of tsetse at the species level . The created main spectra were then compiled as a tsetse database . In order to check the suitability of the created tsetse main spectra for Biotyper-based species identification , the cross-matching status was created after matching them to the entire database . In addition , fresh extractions of the whole insect and the various insect parts were utilized in triplicates to cross-check the efficiency of the established tsetse database . For ruling out possible cross-matching with other fly species , the common house fly Musca domestica ( Diptera: Muscidae; strain Lei ) was also included in the evaluation . Identification was carried out using the Biotyper 3 . 0 software tool , following the manufacturer's recommendation on identification based on the calculated log score values . Values of ≥2 . 0 to 3 . 0 represent probable species level matching , while scores of ≥1 . 7 to 1 . 9 represent probable genus level matching . A score value of <1 . 7 stands for an unreliable identification .
From each tsetse specimen a total of 27 spectra representing biological and technical replicates in the m/z range of 2000–20000 Da were acquired automatically and thus 1620 spectra from whole Glossina species and their body parts A–J . Visual inspection of the spectra revealed a comparable peak pattern of the biological and technical replicates; however , differences in peak intensities were observed for example as shown in figure 1 . At first look , the raw spectrum displayed consistently distinct peak patterns when comparing the two sexes of G . palpalis gambiensis ( figure 2 , samples G/H at m/z 5700 , 7000 and 8000 ) while the three savannah species ( A–F ) and G . brevipalpis ( I/J ) only displayed differences in peak intensity . Occasionally observed differences as seen in the G . pallidipes female ( sample E at 8100 m/z ) appeared inconsistently . However , several peaks showed to be common for Glossina spp . as for instance presented in figure 2 at 5000 m/z . As shown in figure 3 , the raw spectra of different body parts and the entire insects presented varying peak patterns at least in terms of peak intensities . Among the body parts , peak intensities sometimes tended to be lower in some of the leg extracts when compared to entire insects or other parts . To demonstrate the protein composition of whole insects and the different body parts , G . palpalis gambiensis extracts were chosen for protein separation on SDS-PAGE and visualised using a modified Coomassie staining . As shown in figure 4 , the protein separation was carried out from 10 to 200 kDa . The bands out of the extracts of the dissected body parts were clearly observed in the whole insect protein extract lane . However , it should be noted that the peaks in the MALDI spectra were obtained from much smaller peptides ( 2–20 kDa ) . Figure 5 depicts the colour-coded computed composite correlation index [31] displaying the uniqueness of the acquired spectra 1–60 . A CCI value of 0 . 0 ( dark green ) represents incongruency and 1 . 0 ( red ) denotes complete congruency . The CCI was observed between 0 . 68 and 0 . 98 ( individual CCI values are shown in the supplement data table S1 ) . Very few of the spectra sets displayed some deviation among themselves , for e . g . the CCI for G . austeni male head was 0 . 68 . However , this spectra set displayed a complete deviation with other body parts or other species . Despite this shortcoming , the spectra sets appeared to be suitable for the compilation of a reference spectra library . Cross–comparison of the tsetse main spectra with the entire Bruker reference database resulted in only one clear match with a log score value of >2 . 3 , the cut-off value representing the most probable matching at the species-level . Some isolates such as G . austeni female head ( no . 2 ) appeared to resemble G . palpalis gambiensis male head spectra ( no . 56 ) , however , the score value was distinctly lower than the expected matching set . This clearly indicated that these spectra sets could be utilized to establish a database . Following these preliminary investigations , the main spectra library representing the 70 most reproducible peaks was constructed . The cluster analysis of the 10 main spectra of each species is shown in figure 6 for both sexes . Consistent clustering was observed among the extracts of G . brevipalpis , which always stood out as a sister group to the other species regardless of the body part . Furthermore , G . austeni showed inconsistent clustering , neither similar to savannah group tsetse nor to riverine G . palpalis gambiensis as for instance seen in the dendrogram . The created tsetse main spectra were incorporated into the commercial Bruker system and then compared with the whole database following the manufacturer's recommendation . Accordingly , table S2 of the supplementary data describes the matching of tsetse main spectra where log score value 3 . 0 stands for a 100% match and lower matching probabilities were displayed as subsequent hits . The results indicate that the second hit within the acceptable cut-off value of >2 . 0 for some of the body part extracts matched with the correct body part but irrespective of the factors sex and species . This cross matching of body parts was predominantly observed between G . austeni and G . morsitans morsitans and among G . pallidipes and G . palpalis gambiensis . Within the same species , complete deviation was observed in G . austeni female head with its own abdomen and legs , Similarly , G . palpalis gambiensis female head did not match with its legs and thorax . G . palpalis gambiensis male head also displayed complete deviation with G . palpalis gambiensis female head . As shown in table 2 ( detailed identification results are listed in supplementary table S3 ) , the results of fresh sample identification clearly indicate that every body part and sex was correctly matched at the species level ( log score value >2 . 0 ) . Despite the 100% correct identification , within this high confident identification the following score inconsistencies occurred: 58% ( 35/60 ) matched with the correct body part but also with the ones of the opposite sex , 35% ( 21/60 ) matched with the correct sex but with different body parts , 16% ( 10/60 ) matched with a different body part and the opposite sex and 5% ( 3/60 ) even matched with other species . The second best matching hits indicate that about 23% ( 14/60 ) of body parts displayed lower cut-off values ( log score <1 . 7 ) . Among the second best hits , incorrect matching was observed among 13 samples ( 21% ) : body parts of female G . palpalis gambiensis ( thorax , whole and abdomen ) , G . pallidipes ( female thorax ) and G . austeni ( male legs ) . The extracts from Musca domestica resulted in no reliable identification .
To establish a tsetse reference database five laboratory breeds representing epidemiologically important tsetse of the savannah type G . morsitans morsitans , G . pallidipes and G . austeni , a riverine type G . palpalis gambiensis and forest type G . brevipalpis were chosen for this study [2] , [33] , [34] . Earlier attempts on the identification of arthropods by MALDI were carried out after homogenisation of the samples and extraction in a mass spectrometry-compatible buffer system [18] , [20] , [21] . We used a standard formic acid/acetonitrile extraction procedure of microbial cell processing for the protein extraction from tsetse . We introduced an additional step of sonication in order to facilitate the breakage of the chitin shell for a better protein yield . This simple extraction method was chosen to accommodate the field-collected samples that are stored in ethanol and possibly dissected . Flex analysis software revealed that the spectra of the same species appeared to be fairly comparable despite the varying peak intensities . Visual inspection of the spectra revealed differences among the body parts of the same insect . Often , the most intense peaks of body part extracts were not easily observable in the spectra of whole insect extracts . This could be due to the protein ionisation influenced by varying protein compositions/abundances of different body part extracts . Additional evaluation of the protein composition/abundance using SDS-PAGE protein separation revealed the difference in protein bands . However , the bands of the body part extracts were comparable to those of the whole insect but they varied in their intensities . This was also shown among the different sexes of the same species . As the protein separation was carried out in a higher range ( 10 to 200 kDa ) but the MALDI spectra stemmed from a much smaller range of proteins ( 2 to 20 kDa ) , So , a direct correlation among these could not be expected . However , the compositional protein differences among the various body part extracts and the whole insect are clear . This protein compositional difference might attribute to the observed difference among the spectra from different insect body parts . Despite this variation , the technical and biological replicates appeared to influence the peak intensity while the peak pattern was almost comparable . Among the commercially available software tools for species identification , we used Biotyper software that incorporates 4613 main reference spectra of microbial species ( March 2013 ) . The software automatically pre-process the spectra through smoothing and baseline subtraction . The peaks were picked and compared with the reference database . The results were expressed as similarity log score values between 3 . 0 ( complete matching ) to 0 ( complete deviation ) . As a first step of the main spectra creation , the practical relation among the spectra sets was visualised by computing the composite correlation index [32] . A CCI value approaching 1 is considered to be highly significant while zero represents complete deviation . A clear distinction between the spectra sets of different body parts and the whole insect extracts was displayed in the heat-map and its corresponding value . Some of the spectra set displayed signs of deviation , for e . g . G . austeni male head . This might be due to the presence of broader peaks , which did not overlap with the corresponding spectra [32] . The heat-map and CCI values indicated that the spectra sets of different body parts and the whole insect extracts were unique and could be utilised for the creation of a spectra library . Therefore , we generated 60 main spectra for five tsetse species including male and female whole insect extracts and the corresponding body parts . These main spectra were then incorporated in the Bruker database . The main spectra dendrogram was useful for the differentiation of the five species , picturing the similarities and differences of their mass spectra profiles . Clustering of the created tsetse main spectra revealed that they did not follow any distinct pattern with some significant exceptions . A possible explanation could be that higher organisms like insects might not cluster at the species level using MALDI measurements unless they are being standardised . However , G . austeni never clustered clearly with riverine nor savannah species; it seems to share mass spectra patterns with both groups reflecting the uncertainty of their phylogenetic status [35] . Very clearly though was the uniqueness of G . brevipalpis compared to the other species . The sister status deriving from genomic findings [36] could therefore be mirrored in the mass spectrum peaks of G . brevipalpis . As a quality check , tsetse main spectra were cross-identified with the entire database from the manufacturer . All the tsetse main spectra matched with a log score of 3 . 0 , indicating a clear distinction between the species . It also showed the uniqueness of the tsetse mass spectra for entire tsetse as well as every dissected body part . Among the second best matched hits , sex and species appeared to be least important while the body parts across the species matched , especially among G . austeni and G . morsitans morsitans and also in G . pallidipes and G . palpalis gambiensis extracts . The complete deviation of head extracts ( G . austeni female , G . palpalis gambiensis female and male ) indicates special attention when working on species identification of head samples by MALDI . The fresh protein extracts using the same insects resulted in 100% matches with the database . No hits were achieved for similarly processed Musca domestica extracts , indicating the uniqueness of the created reference spectra for tsetse . Among the best hits at the species level , body parts of the same species appeared to be matched correctly but irrespective of the sex . A deviating species in the second hit might be due to the presence of shared metabolic proteins among different tsetse species . The 5% that mismatched completely and the incorrect matching among the second hits indicate that the reference database should be created for more than one body part and of both sexes for reliable identification of insects . The overall results clearly indicate that the success in MALDI-based identification relies on the specific signature from the body parts and the whole insects . While the first hit for these lab breed tsetse appeared to be specific for species , sex and body parts , the second hit indicates that sex is the least reliable feature of MALDI identification . The complete deviation of head extracts with its own other body parts as seen among G . austeni and G . palpalis gambiensis indicate that more than one body part is needed for accurate species identification . We propose the addition of spectra from field-caught tsetse ( whole insects and body parts ) to extend our database for a fast and accurate identification of tsetse .
|
Tsetse flies are confined to tropical Africa and are carriers for trypanosomes , single-celled blood parasites . Through the bite of an infective tsetse , people and animals may contract trypanosomiasis , a degenerative disease leading to death if left untreated . Tsetse control proved effective for disease containment , but data on the flies as tsetse identification are a prerequisite for planning any control intervention . There are 32 generally accepted tsetse species and subspecies . Classical species identification relies on minor morphological differences , often challenging for field workers . In the last decade , Matrix-Assisted Laser Desorption/Ionisation ( MALDI ) has revolutionised microbial species identification . After a simple protein extraction , a laser-induced ionisation takes place . Then , the ions are accelerated in a vacuum tube , and their Time of Flight ( ToF ) to reach the detector is recorded . The protein composition of each organism is unique , and so is their MALDI signature . Comparison of the obtained signature with a database of known organisms enables rapid identification as reliable as genome-based methods . To possibly speed up tsetse diagnostics , we established a MALDI database for the identification of five defined laboratory tsetse breeds . Inclusion of wild-caught tsetse could reinforce the reference database for the identification of tsetse at the species and subspecies level .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"veterinary",
"diseases",
"clinical",
"laboratory",
"sciences",
"diagnostic",
"medicine",
"veterinary",
"parasitology",
"veterinary",
"diagnostics",
"zoonotic",
"diseases",
"livestock",
"care",
"proteomics",
"biology",
"trypanosomiasis",
"veterinary",
"science",
"veterinary",
"medicine"
] |
2013
|
Identification of Tsetse (Glossina spp.) Using Matrix-Assisted Laser Desorption/Ionisation Time of Flight Mass Spectrometry
|
Although the transcription factors IRF-3 and IRF-7 are considered master regulators of type I interferon ( IFN ) induction and IFN stimulated gene ( ISG ) expression , Irf3−/−×Irf7−/− double knockout ( DKO ) myeloid dendritic cells ( mDC ) produce relatively normal levels of IFN-β after viral infection . We generated Irf3−/−×Irf5−/−×Irf7−/− triple knockout ( TKO ) mice to test whether IRF-5 was the source of the residual induction of IFN-β and ISGs in mDCs . In pathogenesis studies with two unrelated positive-sense RNA viruses ( West Nile virus ( WNV ) and murine norovirus ) , TKO mice succumbed at rates greater than DKO mice and equal to or approaching those of mice lacking the type I IFN receptor ( Ifnar−/− ) . In ex vivo studies , after WNV infection or exposure to Toll-like receptor agonists , TKO mDCs failed to produce IFN-β or express ISGs . In contrast , this response was sustained in TKO macrophages following WNV infection . To define IRF-regulated gene signatures , we performed microarray analysis on WNV-infected mDC from wild type ( WT ) , DKO , TKO , or Ifnar−/− mice , as well as from mice lacking the RIG-I like receptor adaptor protein MAVS . Whereas the gene induction pattern in DKO mDC was similar to WT cells , remarkably , almost no ISG induction was detected in TKO or Mavs−/− mDC . The relative equivalence of TKO and Mavs−/− responses suggested that MAVS dominantly regulates ISG induction in mDC . Moreover , we showed that MAVS-dependent induction of ISGs can occur through an IRF-5-dependent yet IRF-3 and IRF-7-independent pathway . Our results establish IRF-3 , -5 , and -7 as the key transcription factors responsible for mediating the type I IFN and ISG response in mDC during WNV infection and suggest a novel signaling link between MAVS and IRF-5 .
The type I interferon ( IFN ) signaling network is an essential component of the innate immune response against viral infections , and also functions to shape adaptive immunity [1]–[4] . Infected cells initiate an antiviral response upon recognition of non-self pathogen-associated molecular patterns ( PAMPs ) , which are detected by host pattern recognition receptors ( PRRs ) [2] , . PRRs that recognize RNA viruses include members of the Toll-like receptor ( TLR3 and TLR7 ) and the RIG-I-like receptor ( RLR; RIG-I and MDA5 ) families . TLRs and RLRs recognize distinct PAMPs in different locations ( extracellular/endosomes and cytoplasm , respectively ) and activate signaling cascades to initiate antiviral and inflammatory responses . TLR3 binds to double-stranded RNA and recruits the adaptor molecule TRIF to activate the kinases TRAF and IKK-ε , which in turn activates the latent transcription factors IRF-3 , IRF-7 , and NF-κB . Single-stranded RNA is recognized by TLR7 , which uses the adaptor molecule MyD88 to activate TRAF and IKK-ε , and subsequently NF-κB- and IRF-7-dependent transcription . RLRs interact with the mitochondria-associated adapter molecule MAVS ( also called IPS-1 , VISA , or CARDIF ) , which signals through the kinases TBK1 and IKK-ε to activate IRF-3 , IRF-7 , and NF-κB and initiate type I IFN production . A canonical model for type I IFN production after RNA virus infection is a two-step positive feedback loop that is regulated by IRF-3 and IRF-7 [9] , [10] . In the first phase , viral sensing by TLRs or RLRs induces nuclear localization of IRF-3 , which in concert with NF-κB and ATF-2/c-Jun stimulates transcription , synthesis , and secretion of IFN-β and IFN-α4 by infected cells . In the second phase , extracellular IFN-β and IFN-α4 bind to the type I IFN receptor ( IFNAR ) , which triggers activation of the JAK-STAT signaling pathway and induction of IFN-stimulated genes ( ISGs ) [11] . ISGs act by a variety of mechanisms to render cells resistant to viral replication [12] , [13] . Although type I IFN signaling is required to activate the full antiviral response , a subset of ISGs is induced directly by IRF-3 [14] , [15] . While IRF-3 is constitutively expressed in many tissues , IRF-7 is an ISG required for the expression of most IFN-α subtypes , and thus a key mediator of the type I IFN amplification loop [2] , [9] , [10] . Certain cells , including plasmacytoid dendritic cells and macrophages , express IRF-7 constitutively , which makes them poised for rapid IFN-α production [16]–[20] . West Nile virus ( WNV ) is a mosquito-transmitted , enveloped , positive-sense RNA virus and member of the Flaviviridae family . Studies in mice with targeted gene deletions have provided insight into mechanisms of innate immune restriction of WNV infection . The type I IFN response is essential to the control of WNV infection , as mice that are defective at producing or responding to IFN cannot control virus replication and succumb rapidly to infection [17] , [21]–[25] . The host antiviral response in vivo is dependent upon both TLR and RLR signaling , as deficiencies in TLRs , RLRs , or their downstream adaptor molecules ( including MyD88 and MAVS ) result in enhanced viral replication and lethality [8] , [22] , [26]–[30] . Recent studies with WNV have suggested that some cell types use non-canonical signaling pathways to induce type I IFN responses . The combined absence of IRF-3 and IRF-7 resulted in uncontrolled WNV replication and more rapid death in Irf3−/−×Irf7−/− double knockout ( DKO ) mice compared to the individual single gene knockout mice [17] , [21] , [22] , [31] . However , even without IRF-3 or IRF-7 , type I IFN was produced by DKO mice infected with WNV or murine cytomegalovirus , albeit at reduced levels compared to wild type mice [22] , [32] . Consistent with the sustained production of type I IFN , lethality in DKO mice infected with WNV or chikungunya virus was not as rapid or complete as in Ifnar−/− mice [22] , [31] , [33] , [34] . Ex vivo experiments with primary myeloid dendritic cells ( mDC ) and macrophages revealed that the IFN-β response after WNV infection was sustained in DKO cells but abrogated in the absence of MAVS [22] , [27] . In contrast , the IFN-β response in neurons and fibroblasts was abolished in the absence of either IRF-3 and IRF-7 or MAVS [22] , [27] . These studies suggested cell type-specific requirements for the transcription factors that induce IFN-β expression in response to WNV infection . To define the transcription factor ( s ) responsible for the IRF-3 and IRF-7-independent production of IFN-β in myeloid cells , we considered another member of the IRF family , IRF-5 . Although IRF-5 was originally identified as an inducer of inflammatory cytokines ( IL-6 and TNF-α ) downstream of TLR-7 and MyD88 signaling , subsequent studies suggested that it could contribute to type I IFN production after viral infection [35]–[37] . In response to Newcastle disease virus ( NDV ) infection , IRF-5 induced overlapping and distinct sets of genes compared to IRF-7 , including stronger induction of IFN-β and the antiviral gene Rsad2 ( Viperin ) [38] . We generated Irf3−/−×Irf5−/−×Irf7−/− triple knockout ( TKO ) mice and found that these mice were highly vulnerable to infection with WNV . The combined loss of IRF-3 , IRF-5 , and IRF-7 largely abrogated type I IFN and ISG expression in mDC , and microarray analysis of WNV-infected mDC revealed a set of genes induced in DKO but not in TKO cells . Because the limited set of genes induced in WNV-infected TKO mDCs was absent in Mavs−/− mDCs , we conclude that the RLR-MAVS signaling pathway dominantly regulates innate immune gene induction in mDCs during WNV infection , and that IRF-3 , IRF-5 , and IRF-7 coordinately mediate this response . Our results establish a new linkage between the IRF-5 and the RLR signaling pathways in induction of the antiviral IFN response .
We hypothesized that IRF-5 might be responsible for the residual IFN-β production in DKO mice , because IRF-5 contributes to Ifnb mRNA expression downstream of the PRR TLR7 and adaptor molecule MyD88 , both of which limit WNV pathogenesis in vivo [28] , [30] , [39] . To test this , we generated Irf3−/−×Irf5−/−×Irf7−/− TKO mice ( Figure S1 ) and defined their response to viral infection . TKO mice were viable , fertile , and produced progeny according to normal Mendelian frequencies ( data not shown ) . We infected WT , DKO , and TKO mice with a virulent WNV strain ( New York 2000 , WNV-NY ) and found that TKO mice succumbed to infection earlier than DKO mice ( mean time to death ( MTD ) : 4 . 0 days versus 5 . 7 , P<0 . 0001 ) . TKO mice died marginally later than Ifnar−/− mice , which do not respond to type I IFN and fail to control WNV replication ( MTD: 4 . 0 days versus 3 . 7 , P<0 . 05 ) ( Figure 1A ) [25] , [31] . Because TKO , DKO and Ifnar−/− mice all succumbed so rapidly to WNV-NY infection , it was difficult to appreciate biologically meaningful differences in susceptibility among the three genotypes . To address this , we infected these mice with an attenuated WNV strain ( Madagascar 1978 , WNV-MAD ) that inefficiently antagonizes JAK/STAT signaling [23] . With this virus , we observed a pronounced increase in mortality of TKO compared to DKO mice ( Figure 1B ) . Whereas 100% of TKO mice succumbed to WNV-MAD infection , only 20% of DKO mice died ( P<0 . 001 ) . TKO mice were equally vulnerable to WNV-MAD infection as Ifnar−/− mice ( P>0 . 05 ) , and no statistical difference in MTD was observed ( 9 . 0 days for TKO versus 8 . 2 days for Ifnar−/− mice , P>0 . 05 ) . Similar results were observed upon infection with murine norovirus ( MNV ) , an unrelated non-enveloped positive-sense RNA virus . TKO mice were more vulnerable to MNV infection than DKO mice , with only 1 of 11 TKO mice surviving , compared to 100% survival for DKO mice ( P<0 . 0001 ) ( Figure 1C ) . However , the TKO mice did not show the same susceptibility as Ifnar−/− mice ( P<0 . 0001 ) , and the MTD was greater in TKO compared to Ifnar−/− mice ( 7 . 8 days versus 5 . 3 days , P<0 . 001 ) . The observation that lethality in TKO mice more closely matched that of Ifnar−/− mice after WNV infection compared to MNV suggests that there may be virus-specific differences in the particular transcription factors responsible for mediating the antiviral response . Overall , the loss of IRF-5 in the setting of an IRF-3 and IRF-7 deficiency renders mice more vulnerable to viral infection and early death , approaching that seen in mice that cannot respond to type I IFN . To understand the basis of the increased susceptibility of TKO mice to viral infection , we infected WT , DKO , TKO , and Ifnar−/− mice with WNV-NY or WNV-MAD and measured viral burden in the draining lymph node , serum , spleen and brain at 2 days ( WNV-NY ) or 6 days ( WNV-MAD ) after infection ( Figure 1D–G ) . Viral infection in TKO mice was similar to that observed in Ifnar−/− mice ( P>0 . 05 ) in all tissues examined , except for the spleen after WNV-MAD infection where titers in TKO mice were greater than in Ifnar−/− mice ( 25-fold , P<0 . 05 ) . After infection with WNV-NY , TKO mice had higher viral loads than DKO mice in the draining lymph node ( 13-fold , P<0 . 01 ) , spleen ( 5-fold , P<0 . 01 ) , and brain ( 9-fold , P<0 . 05 ) . After infection with WNV-MAD , TKO mice had higher viral loads than DKO mice in the serum ( 124-fold , P<0 . 01 ) and spleen ( 169-fold , P<0 . 01 ) . To determine whether the enhanced vulnerability of TKO mice was due to an inability to generate a systemic antiviral response , we measured type I IFN levels in the serum of mice infected with WNV-NY ( 2 days after infection ) or WNV-MAD ( 6 days after infection ) ( Figure 2 ) . Unexpectedly , we detected type I IFN activity in the serum of TKO mice infected with WNV-NY or WNV-MAD , and the amount present was not different from DKO mice ( P>0 . 05 ) . While the serum levels of type I IFN in TKO and DKO mice were diminished compared to WT mice after WNV-NY infection and equivalent to WT after WNV-MAD infection , substantially higher levels of type I IFN were detected in the serum from Ifnar−/− mice ( 29-fold after WNV-NY infection , P<0 . 01; 416-fold after WNV-MAD infection , P<0 . 0001 ) . The high level of type I IFN in Ifnar−/− mice likely is a result of high viral replication in the absence of IFN-mediated antiviral effector functions combined with the absence of IFNAR molecules to bind and internalize type I IFN in the serum . Despite the combined absence of IRF-3 , IRF-5 , and IRF-7 , TKO mice still produced type I IFN after WNV infection , albeit at lower levels in the context of markedly enhanced infection . Myeloid cells retain the ability to produce IFN-β during WNV infection despite the lack of IRF-3 and IRF-7 [22] . To determine if this IFN-dependent antiviral activity was mediated by IRF-5 , we performed multi-step growth analyses with WNV-NY in primary mDC and macrophages derived from WT , DKO , TKO , and Ifnar−/− mice ( Figure 3A and B ) . Viral replication in TKO mDC was greater than in DKO mDC ( 74-fold , P<0 . 0001 ) and equivalent to Ifnar−/− mDC ( P>0 . 05 ) , suggesting that IRF-3 , IRF-5 , and IRF-7 regulate innate immune defense to control WNV replication in mDC . In comparison , TKO macrophages showed little increase in WNV-NY replication compared to DKO cells , and reached lower ( 11-fold , P<0 . 0001 ) peak titers compared to Ifnar−/− macrophages . This suggests that macrophages can restrict WNV-NY infection through an alternative pathway that is independent of IRF-3 , IRF-5 , and IRF-7 , possibly through IRF-1 and/or other transcription factors [40] . To establish whether the disparate ability of TKO mDC and macrophages to control WNV-NY replication was associated with differences in antiviral gene induction , we infected cells and performed western blots to assay expression of ISGs , specifically RIG-I ( DDX58 ) , MDA5 ( IFIH1 ) , STAT1 , IFIT2 ( ISG54 ) and IFIT3 ( ISG49 ) ( Figure 3C and D ) . In TKO mDCs , we did not detect expression of any of the tested ISGs , although these were highly expressed in WNV-infected WT and DKO mDC ( Figure 3C and [22] ) . In contrast , most of these proteins were induced in TKO macrophages , although their expression was delayed compared to WT cells: ISG expression was detected in TKO macrophages at only 48 hours after infection , whereas expression was detected in WT cells within 12 hours of infection . Unlike other ISGs , IFIT3 was not expressed in TKO macrophages even at 48 hours after infection , despite being induced in DKO macrophages [22] . The lack of virus-induced ISG expression in TKO mDC resembled the phenotype observed in cells lacking the RLR-signaling adaptor , MAVS [27] . To further define the ISGs expressed in an IRF-3 , IRF-5 , or IRF-7 dependent manner , we infected mDC and macrophages from WT , DKO , TKO , and Ifnar−/− mice with WNV-NY and measured the induction of Ifnb , Oas1a , Rsad2 , and Cxcl10 mRNA at 24 hours after infection by quantitative reverse transcription polymerase chain reaction ( qRT-PCR ) ( Figure 3E and F ) . These genes were selected as known representatives of different ISG induction pathways . Rsad2 and Cxcl10 can be induced directly by PRR signaling and IRF-3 mediated transcriptional regulation , whereas expression of Oas1a depends more strictly upon IFN-β signaling [14] , [15] , [20] . Consistent with the western blot results , all four genes were induced strongly in WT and DKO mDC , but not in TKO mDC . In contrast , TKO macrophages retained the ability to express Ifnb and the tested ISGs , although the level of gene induction was equivalent to or less than WT cells , even in the context of enhanced viral replication . As expected , Oas1a was not induced in Ifnar−/− cells , although these cells expressed high levels of Rsad2 , Cxcl10 and Ifnb . ISG expression in Ifnar−/− macrophages was especially high , likely secondary to increased viral replication and IRF-3-dependent gene induction . Since TKO mDC failed to induce expression of selected ISGs in response to WNV-NY infection , we tested their capacity to express ISGs in response to other inflammatory stimuli including IFN-β and the PRR agonists poly ( I∶C ) and lipopolysaccharide ( LPS ) ( Figure 4 ) . Although TKO mDC failed to induce Ifnb expression after WNV-NY infection , they retained the ability to respond to its signaling , inducing WT levels of Ifnb , Oas1a , Rsad2 , and Cxcl10 at 24 hours after IFN-β treatment . However , these cells showed an ablated response to poly ( I∶C ) or LPS , with no induction of Ifnb or the tested ISGs . Thus , TKO mDC are defective in transmitting MyD88- and TRIF-dependent signals after PAMP sensing , whereas the JAK/STAT-ISGF3 signaling pathway remains intact . As observed previously , although DKO mDC induced a WT-like pattern of ISGs after WNV infection , they had a diminished response to stimulation by the TLR4 ligand LPS or by poly ( I∶C ) , which is recognized by TLR3 and MDA5 [22] . This suggests that WNV infection activates a broader range of PRRs than poly ( I∶C ) or LPS treatment alone . Analysis of selected ISGs in TKO mDC infected with WNV-NY suggested a profound loss of gene induction , results that also were seen previously in Mavs−/− cells [27] . To evaluate this in greater detail , we performed a microarray analysis to profile gene expression patterns in TKO and Mavs−/− mDC 24 hours after WNV-NY infection at a multiplicity of infection ( MOI ) of 25 . To identify the specific contributions of IRF-5 and type I IFN signaling to the transcriptional response , studies also were performed with WT , DKO , and Ifnar−/− mDCs . The level of WNV infection of the cells used for the microarray was assessed by flow cytometry using an anti-WNV monoclonal antibody ( Figure 5A ) . TKO and Mavs−/− mDC had significantly higher rates of infection compared to WT cells ( P<0 . 05 and P<0 . 01 , respectively ) , whereas infection of DKO and Ifnar−/− mDC surprisingly was not different than WT ( P>0 . 05 ) . Nonetheless , for all genotypes tested , only a fraction ( up to 15% ) of cells stained positive for WNV antigen at 24 hours after infection , suggesting that uninfected cells contributed substantially to the gene induction profile observed in this experiment . Gene induction was measured by comparing WNV-infected samples to mock-infected cells of the same genotype , to control for differential basal expression of some genes . We considered genes to be expressed differentially in response to WNV infection if they exhibited a fold change of ≥1 . 5 and a P-value<0 . 05 . WNV-infected WT mDCs showed a broad transcriptional response , particularly of genes that are induced by PRR and type I IFN signaling . 445 genes were expressed differentially in WNV-infected mDC compared to mock-infected cells ( Table S1 ) . The 50 most upregulated genes ( Figure 5B ) included ISGs with previously described antiviral activity ( Rsad2 , Ifit2 , Ifit3 , Isg15 , Isg20 , and Parp12 ) [13] , [41]–[43] , members of the 2′-5′-oligoadenylate synthetase family ( Oas1g , Oas2 , Oasl1 , and Oasl2 ) [12] , [44] , [45] , components of the PRR/type I IFN ( Ddx58 , Dhx58 , Ifnb1 , Ifna2 , Irf7 , Stat1 , and Stat2 ) and ISG15 ( Isg15 , Ube2l6 , Usp8 ) [12] pathways , as well as nucleotide metabolism factors ( Cmpk2 and Nt5c3 ) . The particular genes upregulated in DKO mDC were similar to those in WT cells , although the magnitude of induction was lower in DKO cells , consistent with previous observations [22] . In contrast , a restricted set of 22 genes was expressed differentially in WNV-infected Ifnar−/− mDCs ( Figure 5B and Table S2 ) . Remarkably few genes were expressed differentially in either TKO or Mavs−/− mDC upon WNV-NY infection , suggesting that the RLR signaling pathway is critical for initiating the type I IFN and antiviral responses in this cell type . To validate the results of the microarray analysis , we performed qRT-PCR with the same RNA samples that were used for transcriptional profiling ( Figure 5C ) and measured the expression of Cxcl10 , Rsad2 , Ifit2 , Ifnb , Ddx58 , Ccl5 , Ifitm3 , and Ccl2 . The induction pattern measured by qRT-PCR corroborated the microarray results . These eight genes ( listed above in order of relative expression level ) were induced in WT and DKO cells but not in TKO or Mavs−/− cells . Consistent with the patterns observed by microarray , Cxcl10 , Rsad2 , Ifit2 , Ifnb , and Ccl5 were induced in Ifnar−/− cells ( i . e . , are IFN-independent ) , whereas Ddx58 , Ifitm3 , and Ccl2 were not ( i . e . , are IFN-dependent ) . Ifit1 ( ISG56 ) is an ISG that is highly upregulated upon WNV infection [17] , [21] , [27] , [46]–[49] , thus its absence from the infection-induced bioset was unexpected . Upon further analysis by qRT-PCR , we found that Ifit1 was induced to high levels in infected WT , DKO , and Ifnar−/− mDC but not TKO or Mavs−/− cells . This quality control assessment reveals that the single Ifit1 probe on our microarray chip was defective , and that Ifit1 expression is induced in Ifnar−/− cells after WNV infection . To identify genes whose expression was dependent strictly upon IRF-5 and MAVS , we considered those upregulated in WT but not in Mavs−/− cells ( MAVS-dependent ) or in WT and DKO but not in TKO cells ( IRF-5 dependent ) . Since TKO and Mavs−/− mDC failed to produce IFN-β in response to WNV infection ( Figure 3 and [27] ) , we stratified our analysis to consider only genes that were upregulated in Ifnar−/− mDC , so as to exclude those whose differential expression might be secondary to the lack of IFN signaling in Mavs−/− and TKO cells . The IFN-independent set of genes ( Figure 6A and Table S2 and S3 ) included Ifnb1 , Rsad2 , Isg15 , Cxcl10 , Ifit2 , and Ifit3 , all of which are induced by IRF-3 without a requirement for IFNAR-mediated signaling [14] , [15] . Further analysis revealed that IFN-independent genes included cytokines ( Ifnb1 , Tnf , Il6 ) , chemokines ( Cxcl10 , Ccl5 , Ccrl2 ) , antiviral restriction factors ( Rsad2 , Isg15 , Ifit2 , Ifit3 ) , and components of the unfolded protein response ( Ppp1r15a ( GADD34 ) , Ddit3 ( CHOP , GADD153 ) , Chac1 ) . To corroborate this analysis , we measured the expression of Trib3 , Ddit3 , Ppp1r15a , Rgs1 , Nfkbiz , and Chac1 by qRT-PCR using the same RNA samples used for the microarray ( Figure 6B ) . We confirmed that three of these genes were upregulated in WNV-infected TKO mDC ( Trib3 , Ddit3 , and Gadd45a ) ( Figure 6C ) . The qRT-PCR data did however , yield some differences: ( a ) Trib3 induction was not detected in Mavs−/− mDC by qRT-PCR; ( b ) Ddit3 was upregulated in a MAVS-independent manner; ( c ) Rgs1 and Nfkbiz were not upregulated in TKO cells; ( d ) while Ppp1r15a was upregulated in Ifnar−/− mDC , it also was induced in DKO mDC; and ( e ) by qRT-PCR we failed to detect expression of Chac1 in mock- or WNV-infected mDC of any genotype , although it was induced in WNV-infected cortical neurons ( data not shown ) . The absence of gene induction in TKO mDC compared to DKO cells could reflect a direct role for IRF-5 in ISG induction or an indirect effect of the loss of IFN-β production in TKO mDC . To test this , we inhibited type I IFN signaling in DKO cells using an IFNAR-blocking monoclonal antibody ( MAR1-5A3 , [50] ) and used qRT-PCR to measure gene induction in response to WNV-NY infection ( Figure 6D ) . As expected , the IFNAR-blocking antibody prevented induction of Oas1a , a known IFN-dependent ISG [15] , but did not impair induction of Ifnb . Ccl5 and Tnf were induced too weakly to observe differences between the IFNAR-blocking and control MAbs . However , the IFNAR-blocking antibody abolished induction of Cxcl10 , Rsad2 , Ifit1 , and Ifit2 , even though these genes are considered to be IFN-independent [14] , [15] and were induced in Ifnar−/− mDC ( Figure 5C ) . Collectively , these results suggest that IRF-5 contributes to the induction of IFN-β expression after WNV infection in mDC , but does not induce ISG expression directly . To further define the contribution of IRF-5 to IFN and ISG induction in mDC , we infected WT , Irf5−/− , and DKO mDC with WNV ( Figure 6E ) and WT , Irf5−/− , DKO , and TKO cells with Sendai virus ( SeV ) , a negative sense RNA paramyxovirus ( Figure 6F ) and measured gene expression by qRT-PCR . We found no change in the induction of Ifnb , Oas1a , Rsad2 , or Cxcl10 in Irf5−/− mDC compared to WT cells ( P>0 . 05 ) , indicating that loss of IRF-5 alone in mDCs is not sufficient to impact the antiviral response , analogous to results seen with IRF-3 [21] . Consistent with this observation , we observed no significant difference in WNV-NY replication between Irf5−/− and WT mDC ( P>0 . 05 ) ( Figure 6G ) . Although DKO mDC retained intact IFN and ISG responses after WNV infection , this pattern surprisingly was not observed following SeV infection: the induced expression of several ISGs ( Oas1a , Rsad2 , and Cxcl10 ) was lost in both DKO and TKO mDC . While our results with DKO and TKO cells after WNV infection establish that IRF-5 contributes to the type I IFN response in mDCs , the critical nature of the IFN induction pathways in these key sentinel cells may have resulted in the maintenance of redundant signaling pathways to sustain antiviral gene programs . Indeed , the distinct ISG induction phenotypes after WNV and SeV infection in DKO and TKO mDCs suggest that activation of these parallel pathways may differ among diverse viruses . The similar gene induction profiles observed between TKO and Mavs−/− mDC by microarray and qRT-PCR suggested a functional interaction between IRF-5 and MAVS . To test this hypothesis , we transfected WT , DKO , and TKO immortalized mouse embryonic fibroblasts ( MEFs ) with plasmids encoding myc-tagged forms of a constitutively active RIG-I ( N-RIG ) and/or IRF-5 . Ectopic expression of N-RIG and IRF-5 was detected in MEFs 24 hours after transfection by western blotting ( Figure 7A ) and qRT-PCR ( data not shown ) . As expected , we observed increased expression of ISGs ( e . g . , Rsad2 , Ifit1 , and Oas1a ) in WT MEFs transfected with N-RIG compared to untransfected cells ( Figure 7B–D ) . Transfection of N-RIG alone in DKO cells failed to induce these ISGs , suggesting that endogenous IRF-5 in MEFs is not adequately expressed or activated to induce ISGs after a MAVS-dependent signal; these results agree with prior studies showing that the combined loss of IRF-3 and IRF-7 in MEFs abolished the ISG response after WNV infection [22] , [27] . In comparison , co-transfection of N-RIG and IRF-5 together but not IRF-5 alone enhanced ISG induction in DKO and TKO MEFs . Thus , MAVS-dependent induction of ISGs can occur through an IRF-5-dependent yet IRF-3 and IRF-7-independent pathway .
In the present study , we generated Irf3−/−×Irf5−/−×Irf7−/− TKO mice to establish that these three IRF family transcription factors coordinately regulate IFN-β production and ISG expression in mDC . We found that antiviral gene induction was ablated almost entirely in mDC from TKO or Mavs−/− mice , suggesting a dominant role for MAVS in initiating the antiviral response and pointing to a novel signaling interaction between IRF-5 and the RLR signaling pathway . As TKO mice succumbed to WNV infection with similar kinetics compared to Ifnar−/− mice , we expected they would be completely defective at producing type I IFN . Nonetheless , we detected type I IFN activity in the serum of infected TKO mice , suggesting that some cells must produce type I IFN by a pathway that is independent of IRF-3 , IRF-5 , and IRF-7 . Macrophages or related cells ( e . g . , inflammatory monocytes ) may be one source of this residual type I IFN in vivo , as TKO macrophages cultured ex vivo expressed Ifnb as well as a subset of ISGs in response to WNV infection . Type I IFN induction in TKO macrophages could be mediated in part by IRF-1 , which regulates expression of antiviral genes independently of type I IFN in the context of several other viral infections [13] , [51] , [52] . Consistent with this , Irf1−/− macrophages supported enhanced WNV replication compared to WT controls [40] , and viral replication in TKO macrophages did not phenocopy Ifnar−/− cells . Nonetheless , IRF-1 was not sufficient to induce the full complement of ISGs in macrophages , as Ifnb and ISG expression in TKO macrophages was diminished and delayed compared to WT cells . Furthermore , IFIT3 was not expressed in TKO macrophages , although it was sustained in DKO cells [22] . It remains unclear whether the genes upregulated in TKO macrophages were induced by IRF-1 directly , by another transcription factor , or downstream of IFN-β production by these cells . We measured ISG induction in infected mDC to determine whether a lack of antiviral effector gene expression explained the failure of TKO mice and mDC to control WNV replication . In our experiments , fewer than 15% of mDC were infected at 24 hours , even when a high MOI of 25 was used . Increasing the MOI to 100 achieved only marginally higher rates of infection ( data not shown ) and was not practical for the scale of the microarray experiments . Sorting infected cells by flow cytometry prior to transcriptional profiling analysis was not feasible as infected cells must be permeabilized to detect intracellular WNV antigens and recombinant WNV expressing green fluorescent protein are attenuated and/or unstable [53]–[55] . In our microarray studies , uninfected cells likely contributed substantially to the ISG expression signatures observed . Indeed , few genes were induced in WNV-infected TKO or Mavs−/− mDC , even though these cells would be expected to upregulate genes associated with cell stress , survival , and metabolism in response to replication by a cytopathic virus . Some components of the unfolded protein response , including Ddit3 and Gadd45a , were upregulated in infected TKO mDC; additional genes likely were induced in infected cells but may have been below the statistical cutoffs used in our analysis due to dilution of the transcripts in a large pool of mRNA from uninfected cells . Viral infection induces the expression of ISGs both directly ( by IRF-3 after PAMP detection and PRR signaling ) and indirectly ( by IFN-β production and IFNAR signaling ) , the latter occurring in both infected and uninfected cells . Given the large proportion of uninfected cells , we would expect genes induced by IFNAR signaling to predominate . Indeed , only a small subset of genes was induced after WNV infection of Ifnar−/− mDC ( 22 genes , compared to 445 in WT mDC ) . This may reflect the relatively low infection rates , an inherent inefficiency of IFNAR-independent gene induction pathways , or viral countermeasures that antagonize the type I IFN response in highly infected cells [56] . Of the 22 genes induced in WNV-infected Ifnar−/− mDC , several ( Ifnb , Cxcl10 , Rsad2 , Ifit1 , and Ifit2 ) have direct or indirect antiviral activity against WNV [13] , [24] , [41] , [42] , [57]–[59] and are induced directly by IRF-3 [14] , [15] . Other genes induced in WNV-infected Ifnar−/− mDC included components of the unfolded protein response , such as Ddit3 and Ppp1r15a . Ddit3 ( CHOP ) has been shown to promote expression of Ppp1r15a ( Gadd34 ) and Trib3 [60]–[62] , two IFN-independent induced genes detected in our microarray analysis . While induction of these genes may represent a response to the cellular stress caused by viral infection , the unfolded protein response also constitutes a cellular defense that limits replication of diverse viruses , including WNV [60] , [63] , [64] . DDIT3 inhibits WNV replication , and WNV may induce expression of Ppp1r15a to reverse DDIT3-mediated translational inhibition [60] . In contrast , PPP1R15A is required for IFN-β production and contributes to controlling replication of chikungunya virus [65] . Although global gene induction in response to WNV infection has been reported previously [46]–[49] , [66] , [67] , our results represent the first such analysis in DCs , which are a sentinel cell type coordinating the innate and adaptive antiviral immune responses , as well as among the first cells infected following a mosquito bite [8] , [68] . Some of the genes we identified in mDCs also were detected in microarray analyses of WNV-infected MEFs [46] , human kidney epithelial cells [48] , or human retinal pigmented epithelium [47] . Induction of these genes ( e . g . , Rsad2 , Ifit2 , Isg15 , Isg20 , and Stat1 ) thus does not depend on cell type-specific transcription factors . Other WNV-induced genes , however , may be specific to DCs or restricted cell types . As an example , the chemokine Cxcl10 was one of the most highly induced genes in our analysis , yet it was induced at much lower levels or not at all in fibroblasts and epithelial cells [46]–[48] . CXCL10 contributes to clearance of WNV infection from the CNS by recruiting effector T cells , and is the dominant chemokine secreted by neurons after WNV infection [57] . Only one of the 22 genes differentially expressed in Ifnar−/− mDC , Ddit3 , was induced in Mavs−/− mDC , suggesting that the IFN-independent induction signal is conveyed almost entirely by MAVS . Since Mavs−/− mDC failed to produce IFN-β , we surmise that both type I IFN-dependent and -independent pathways of ISG induction are abrogated in these cells . This conclusion agrees with earlier studies on induction of selected sets of genes in Mavs−/− mDC infected with WNV or rabies virus [27] , [69] . Although Mavs−/− cells should retain TLR-mediated antiviral gene induction pathways ( which signal through TRIF and MyD88 ) , we observed almost no ISG induction in Mavs−/− mDC after WNV infection . Thus , RLRs likely are the dominant PRRs that sense WNV infection in mDC; these results are consistent with the essentially intact antiviral responses reported in WNV-infected Tlr3−/− and Myd88−/− mDC [26] , [28] . Although our microarray and qRT-PCR analyses identified 16 genes that were differentially expressed in WNV-infected Ifnar−/− and DKO but not TKO mDC , when gene expression was analyzed from WNV-infected DKO cells that were treated with an antibody blocking type I IFN signaling , only Ifnb gene induction was sustained . These data suggest that in the absence of IRF-3 and IRF-7 , IRF-5 is sufficient to induce IFN-β production in response to WNV infection , but unlike IRF-3 [14] , [15] , does not induce ISGs directly ( Figure 8 ) . Although IRF-5 has been suggested to promote IFN-independent expression of some ISGs including Pkr and Isg20 in NDV-infected cells [38] , IRF-3 may have contributed to these responses . The observed anti-WNV response in DKO mDC likely results from IRF-5-dependent IFN-β production , and the uncontrolled viral replication in TKO mDC is secondary to a lack of IFN-β and resultant absence of ISG induction . This model suggests that cell types having ancillary pathways for IFN-β induction ( such as IRF-1 in macrophages ) can mount antiviral responses even in the absence of IRF-3 , IRF-5 , and IRF-7 . We did not anticipate that the Mavs−/− and TKO mDC would phenocopy one other with respect to ISG induction , since IRF-5 has not been previously implicated in the RLR signaling pathway [35]–[37] . IRF-5 originally was described as an inducer of pro-inflammatory cytokines ( e . g . , IL-6 and TNF-α ) but subsequently was suggested to contribute to the type I IFN antiviral response . Irf5−/− mice have increased susceptibility to viral infections , slightly reduced levels of type I IFN in serum , and more significantly reduced levels of pro-inflammatory cytokines [35] , [37] . IRF-5 expression and antiviral activity , however , appears restricted to a limited set of cell types , including monocytes and DCs [35] , [39] , [70] . Thus , a relative absence of IRF-5 expression in fibroblasts and neurons may explain the observation that type I IFN induction after WNV infection in these cell types is abolished by the combined deletion of IRF-3 and IRF-7 [22] . However , the ability of alternate IRFs to compensate for IRF-3 and IRF-7 in fibroblasts also may depend on the particular viral stimulus , as type I IFN production was essentially absent in DKO fibroblasts infected with WNV , herpes simplex virus , vesicular stomatitis virus , or encephalomyocarditis virus [19] , [22] , but low-level production of Ifnb and Ifna2 mRNA was sustained in DKO fibroblasts infected with chikungunya virus [33] . IRF-5 preferentially stimulates the IFN-β and IFN-α4 promoters , rather than other IFN-α subtypes , which also suggests that it contributes to the primary type I IFN response , prior to amplification via autocrine and paracrine signaling [35] . The IFN-α subtypes induced in IRF-5-expressing cells vary from those induced in IRF-7-expressing cells , suggesting that the IRF expression patterns within a cell modulate the breadth of the type I IFN response [70] . Although MAVS previously was known to induce IFN-β production via IRF-3 and IRF-7 , our experiments suggest that RLR signaling also activates IRF-5 to induce IFN-β production in mDC; the subcellular location where this occurs ( e . g . , mitochondrion ) and through what signaling intermediates remains unknown . A recent study suggested that activation of RLR signaling acts to inhibit induction of inflammatory cytokines by IRF-5 [71]; although the net result was different , this study is consistent with our observation of a functional interaction between IRF-5 and MAVS and with a prior proteomic study demonstrating a physical interaction between these two proteins [72] . Future studies will be required to delineate the mechanistic and functional intermediates that link and regulate the IRF-5 and RLR signaling pathways .
The WNV-NY strain ( 3000 . 0259 ) was isolated in New York in 2000 and passaged once in C6/36 Aedes albopictus cells to generate a virus stock that was used in all experiments except for the microarray analysis [73] , [74] . For the microarray studies , mDCs were infected in the Früh laboratory with the WNV New York 1999 strain that was propagated in C6/36 cells [75] . The attenuated strain WNV-MAD was amplified in Vero cells and has been previously described [23] . MNV strain MNV1 . CW3 [76] was propagated in RAW 264 . 7 cells ( ATCC ) and a concentrated stock was prepared as previously described [77] . The SeV virus strain Fushimi was propagated in chicken embryos and provided by D . Lenschow and M . Holtzman ( Washington University , St Louis , MO ) . 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 at the Washington University School of Medicine ( Assurance Number: A3381-01 ) . Dissections and footpad injections were performed under anesthesia that was induced and maintained with ketamine hydrochloride and xylazine , and all efforts were made to minimize suffering . All mice used were on an inbred C57BL/6 background . WT mice were commercially obtained ( Jackson Laboratories ) . Irf3−/−×Irf7−/− DKO , Irf5−/− , and Ifnar−/− mice have been reported previously [22] , [31] , [36] . Irf3−/−×Irf5−/−×Irf7−/− TKO mice were generated by crossing DKO and Irf5−/− mice . Irf5−/− and TKO mice were genotyped for a mutation in the Dock2 gene , which can arise spontaneously in some Irf5−/− mice [78]; none of the TKO mice had homozygous mutations in Dock2 . Mavs−/− mice were generated directly from C57BL/6 embryonic stem cells [34] . All deficient mice were bred in the animal facilities of the Washington University School of Medicine and genotyped prior to experimentation . For WNV infections , 102 PFU was diluted in Hank's Balanced Salt Solution supplemented with 1% heat-inactivated fetal bovine serum and 8 to 12 week-old mice were inoculated by footpad injection in a volume of 50 µl . For MNV infections , 7 to 8 week-old mice were inoculated orally with 3×107 PFU in 25 µl of PBS and monitored for survival for 21 days . To monitor viral spread in vivo , mice were infected with 102 PFU of virus and sacrificed at 2 days after infection ( WNV-NY ) or 6 days after infection ( WNV-MAD ) . After extensive perfusion with PBS , organs were harvested , weighed , homogenized and virus was titered by plaque assay on BHK21-15 cells [74] . Viral burden in serum and inguinal lymph node was measured using fluorogenic qRT-PCR using primers and probes to WNV-NY or WNV-MAD envelope gene sequences ( Table S4 ) . Viral RNA in the lymph node was normalized to Gapdh levels in tissue samples . Viral RNA from serum was isolated using a Viral RNA Mini Kit ( Qiagen ) . Total RNA from lymph nodes was extracted using the E . Z . N . A . total RNA kit ( Omega Bio-tek ) and DNase-treated to remove genomic DNA . Quantitative RT-PCR was performed using One-Step RT-PCR Master Mix and a 7500 Fast Real-Time PCR System ( Applied Biosystems ) . Levels of biologically active type I IFN in serum were determined using an encephalomyocarditis virus L929 cytopathic effect bioassay as described [79] . The amount of type I IFN per ml of serum was calculated from a standard curve using IFN-β ( PBL InterferonSource ) and adjusted for the background inhibitory activity of naïve serum ( approximately 0 . 1 IU/ml ) . The inhibitory activity of naïve serum was type I IFN-independent because it was acid labile but resistant to treatment with heat ( 56°C ) or the IFNAR-blocking antibody MAR1-5A3 [17] , [50] . Macrophage and mDC cultures were generated as described previously [79] . Briefly , bone marrow was isolated from WT , DKO , TKO , Irf5−/− , or Ifnar−/− mice and cultured for seven days in the presence of 40 ng/ml M-CSF ( PeproTech ) to generate macrophages or with 20 ng/ml GM-CSF and 20 ng/ml IL-4 ( PeproTech ) to produce mDC . Multi-step virus growth analysis was performed after infection at a MOI of 0 . 01 for macrophages or 0 . 001 for mDCs . Supernatants were titered by focus-forming assay on Vero cells using humanized E16 anti-WNV MAb as the detection antibody [80] , horseradish peroxidase conjugated anti-human IgG ( Sigma ) , and True Blue Peroxidase Substrate ( KPL ) . For western blotting , cells were infected at an MOI of 1 . For measurement of ISG induction by qRT-PCR , cells were infected at an MOI of 0 . 1 . To block signaling by type I IFN , DKO cells were treated with 25 µg/ml of the IFNAR-blocking MAb MAR1-5A3 for one hour prior to infection . A non-binding MAb against human IFN-γ receptor ( GiR-208 ) was used as an isotype control [50] . Bone marrow cells were cultured in RPMI supplemented with 10% fetal bovine serum , penicillin/streptomycin , L-glutamine , non-essential amino acids , 55 µM β-mercaptoethanol and 20 ng/ml recombinant mouse GM-CSF ( eBioscience ) for six days in non-tissue culture treated plates . GM-CSF was replenished after two days and non-adherent cells were sub-cultured after 4 days . Sub-cultured cells were infected at an MOI of 25 with WNV-NY . Total RNA was harvested at 0 , 6 , 12 , and 24 hours post-infection with an RNeasy Mini Kit ( Qiagen ) . RNA was treated with DNase prior to cDNA generation . Gene expression was assayed on Illumina microarray chips . Microarray datasets were processed by quantile normalization and annotated using the illuminaMousev2 . db R package version 1 . 10 . 0 . Data were assessed by linear modeling with the limma package [81] . Differentially expressed genes were identified as those with at least a 1 . 5-fold change as compared to controls and a P-value<0 . 05 without correction for false discovery . WNV-infected samples were first compared with mock-infected controls . Microarray data have been deposited in GeoArchive , series number GSE42232 . MEFs prepared from WT , DKO , or TKO mice were immortalized after transfection with the plasmid pSV2 , which encodes for the large T antigen of SV40 . MEFs were transfected using Lipofectamine 2000 ( Invitrogen ) with plasmids expressing myc-tagged forms of murine IRF-5 ( Origene ) or residues 1–229 of human RIG-I ( N-RIG ) [82] . Cells were lysed 24 hours post-transfection and analyzed by qRT-PCR and western blotting . Macrophages and mDC were lysed in RIPA buffer ( 10 mM Tris , 150 mM NaCl , 0 . 02% sodium azide , 1% sodium deoxycholate , 1% Triton X-100 , 0 . 1% SDS , pH 7 . 4 ) , with protease inhibitors ( Sigma ) . Samples ( 20 µg ) were resolved by electrophoresis on 10% SDS-polyacrylamide gels . MEFs were lysed in RIPA buffer and lysates were separated by electrophoresis on 4–12% SDS-polyacrylamide gels . Following transfer of proteins , membranes were blocked with 5% non-fat dried milk and probed with the following panel of primary antibodies: rabbit anti-IFIT2 and -IFIT3 ( provided by Dr . G . Sen , [83] ) ; rabbit anti-RIG-I and anti-MDA5 ( IBL ) ; mouse anti-tubulin ( Sigma ) ; rabbit anti-GAPDH ( Santa Cruz ) ; rabbit anti-STAT1 ( Cell Signaling ) ; goat-anti WNV NS3 ( R&D Systems ) ; mouse anti-myc ( Santa Cruz ) . Western blots were incubated with peroxidase-conjugated secondary antibodies ( Jackson Immunoresearch and Sigma ) and visualized using ECL reagents ( Amersham Biosciences and Pierce ) . mDCs were treated for 24 hours with 500 IU/ml of IFN-β ( PBL Interferon Source ) , 50 µg/ml of poly ( I∶C ) ( InvivoGen ) , or 5 µg/ml of LPS ( List Biological Laboratories ) . Macrophages and mDC were infected with WNV-NY at an MOI 0 . 1 for 24 hours . MEFs were harvested 24 hours after transfection . Total RNA was extracted using the E . Z . N . A . total RNA kit ( Omega Bio-tek ) or RNeasy kit ( Qiagen ) and treated with DNase . Fluorogenic qRT-PCR was performed using One-Step RT-PCR Master Mix and a 7500 Fast Real-Time PCR System ( Applied Biosystems ) with the indicated Taqman primers and probes ( Table S4 ) . Gene induction was normalized to Gapdh levels and expressed on a log2 scale as fold increase over mock according to the ΔΔCt method [84] . Data were analyzed with GraphPad Prism software . Viral burdens were compared using the Mann-Whitney test . Serum type I IFN levels , viral growth curves and qRT-PCR were compared using a 2-way ANOVA . Kaplan-Meier survival curves were analyzed by the log rank test and mean times to death were compared by Student's T-test .
|
Host pathogen sensors , including those of the Toll-like receptor and RIG-I like receptor ( RLR ) families , detect viral infection in cells . Signaling through these receptors triggers expression of type I interferon ( IFN ) and IFN-stimulated genes ( ISGs ) , in part through the IRF family of transcription factors . Previous studies with West Nile virus ( WNV ) showed that IRF-3 and IRF-7 control IFN expression in fibroblasts and neurons , whereas macrophages and myeloid dendritic cells ( mDC ) retained the ability to induce IFN-β without IRF-3 and IRF-7 . In the current study , we generated Irf3−/−×Irf5−/−×Irf7−/− ( TKO ) mice to characterize the contributions of specific IRF transcription factors to IFN and ISG induction in response to WNV infection in cells and in mice . We found that induction of IFN and ISGs was largely abolished in TKO mDC , but sustained in TKO macrophages . Because IFN and ISG induction also was absent in mDC lacking MAVS , a key mediator of RLR signaling , our results suggest a novel signaling link between IRF-5 and MAVS . This study establishes the molecular pathways responsible for IFN induction in mDC and suggests a cross-talk between IRF-5 and RLR signaling pathways .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"immunity",
"to",
"infections",
"immunology",
"microbiology",
"immune",
"defense",
"infectious",
"disease",
"control",
"animal",
"models",
"of",
"infection",
"infectious",
"diseases",
"biology",
"immune",
"response",
"west",
"nile",
"fever",
"immunity",
"virology",
"innate",
"immunity",
"microbial",
"control",
"viral",
"diseases"
] |
2013
|
IRF-3, IRF-5, and IRF-7 Coordinately Regulate the Type I IFN Response in Myeloid Dendritic Cells Downstream of MAVS Signaling
|
Nicotinic acetylcholine receptors ( nAChRs ) are widely expressed throughout the central nervous system and modulate neuronal function in most mammalian brain structures . The contribution of defined nAChR subunits to a specific behavior is thus difficult to assess . Mice deleted for ß2-containing nAChRs ( ß2−/− ) have been shown to be hyperactive in an open-field paradigm , without determining the origin of this hyperactivity . We here develop a quantitative description of mouse behavior in the open field based upon first order Markov and variable length Markov chain analysis focusing on the time-organized sequence that behaviors are composed of . This description reveals that this hyperactivity is the consequence of the absence of specific inactive states or “stops” . These stops are associated with a scanning of the environment in wild-type mice ( WT ) , and they affect the way that animals organize their sequence of behaviors when compared with stops without scanning . They characterize a specific “decision moment” that is reduced in ß2−/− mutant mice , suggesting an important role of ß2-nAChRs in the strategy used by animals to explore an environment and collect information in order to organize their behavior . This integrated analysis of the displacement of an animal in a simple environment offers new insights , specifically into the contribution of nAChRs to higher brain functions and more generally into the principles that organize sequences of behaviors in animals .
nAChRs are well-characterized transmembrane allosteric oligomers composed of five identical ( homopentamers ) or different ( heteropentamers ) subunits [1] . Nine different subunits are widely expressed in the mammalian brain , modulating neurotransmitter release , neuronal excitability and activity dependent plasticity in most , if not all , mammalian brain structures [2] , [3] . The elementary mechanisms of nAChRs functions are investigated in great details , yet important issues relevant for the role of nAChRs at the higher level , have received less attention . The need to fill this gap is reinforced by nAChR participation in a diverse array of neuropathologies , including Alzheimer's disease , Parkinson's disease , schizophrenia , epilepsy and Attention-deficit hyperactivity disorder . The complex nature of all these disorders underlines the nicotinic influences over neuronal circuits involved in attention , motivation and cognition [2] , [3] . The issue then becomes how to tackle this problem in mouse models that allow pharmacological and genetic manipulations , but for which “psychological” processes must be inferred from observable behaviors . Mice deleted for ß2-subunit containing nAChR ( ß2−/− ) have been the first nicotinic receptor mutant to be characterized , and found to exhibit more rigid behavior and less behavioral flexibility than wild-type ( WT ) animals [4] . Overall , these experiments suggest that ß2−/− mice reduce the time allocated to explore a novel environment [4] , [5] . Lentiviral reexpression techniques indicate that this phenotype is linked to the expression of ß2*-nAChRs in the ventral tegmental area [6] , [7] and in the Substantia Nigra [8] . ß2−/− mice were shown to be hyperactive in an open-field paradigm , with a reduced movement at low speed , and consequently an increased movement at high speed . Hyperactivity in an open field is often used as a general and non-specific term characterizing experimental conditions where animals show either an increased amount of displacement and related locomotor behaviors , or changes in the frequency of specific motor acts [9] . Increased locomotor activity in an open field can reflect different processing and alterations in the organization of behavior [9] . A complete description of hyperactivity then requires to study duration and temporal patterning ( i . e . the sequence ) of behavioral acts . In this paper , we address the problem of tracing , by analyzing temporal organization of movement , mouse cognitive and/or decision making behavior that can account for mouse hyperactivity in the open-field . Open-field behaviors have been used to study forced exploration of a new environment . It has been shown that it involves both exploratory and stress/fear components [10]–[13] . Furthermore , kinematic features based on instantaneous speed and location have been used to demonstrate that rat and mouse trajectories are far from random [14] , [15] , and that animals can stop more frequently in specific locations of the field that structure their trajectory [16] , [17] . Here , we focus on the analysis of the behavioral sequence , namely the time-organized sequence of patterns that composes the behavior . Considering a sequence of acts , a question would be whether information contained in the structure of this sequence and the presence of specific associations between acts reflects decision-making behavior and can be used to assess alterations of this process . We developed further the method already successfully applied to detect modifications of locomotor behavior caused by mutations in ß2−/− mice [4] , [6] , or in goldfish [18] . The principle of the method is to decompose animal trajectories into a combination of discrete units extracted by applying a threshold to continuous variables . We show that the use of a variable-length Markov model [19] to analyze the sequence of symbols allows to unravel significant alterations in the way ß2 mutant mice organize their behavior , and use “stops” to explore their environment .
Both WT and ß2−/− mice were active in the open-field . They exhibited movements along the wall , sequences of trajectories in the middle of the field ( Figure 1A ) , and alternation between locomotor progression and periods of slow movements . This allowed us to describe locomotor activity in terms of a sequence of four states {PI , PA , CI , CA} ( Figure 1B and 1C ) . ß2−/− mice have been shown to be hyperactive in the open-field ( Granon et al 2003 , Avale et al , 2008 ) , with a distance traveled during 30 min being 1 . 25 times longer in KO compared to WT mice ( Figure 2A , Δ = 34 . 57 m ) . This hyperactivity was reflected in the time spent in an inactive or active state with a decreased time in the inactive state in mutant mice ( Figure 2B ) . The relation between the distance traveled and the duration of the different states were however not different in the two strains . For both strains , the distance traveled during active or inactive states was different , but both exhibited a linear relationship with the duration of a given event ( μ = 0 . 113 and 0 . 117 in active phase for wt and ß2−/− mice , and μ = 0 . 02 and 0 . 023 in inactive phase ) . These relationships tended to break down for long events , but were not different in WT ( Figure 1C , left ) and in ß2−/− mice ( Figure 1C , right ) . The distance traveled was then roughly reflected in the time spent in inactive or active states . These results suggest that higher locomotor activity in ß2−/− mice is not due to a modification of the velocity distribution ( either in the active or inactive phase ) , but rather to a significant change in the organization of the behavior . A change in the time spent in inactive states does not give any insight into the modification of the temporal structure of behaviors . Analysis of transition frequencies and conditional probabilities between different states of the animal were then carried out ( Figure 3A1 ) . Using only four states {PI , PA , CI , CA} did not allow to build a first-order Markov description of the sequence of states . Indeed , when checking for all possible combinations of states X , Y , Z whether P ( X|YZ ) = P ( X|Y ) was satisfied , revealed that the probability of states X after Y = PA did not depend only on the present state PA , but also on the previous one Z ( Figure 3A2 ) . In order to obtain a first order Markov dynamics , PA symbols had to be differentiated into peripheral movement that follows central movement ( CA ) , and peripheral movement that follows inactivity in the periphery ( PI ) . They will be designated by the symbols PAc and PAp , respectively . Using the five symbols {PAc , PAp , PI , CA , CI} allowed to describe open-field activity by a first-order process ( Figure 3A3 ) . This implies that , with such a state description , the animal movement depends only on the preceding state , suggesting a very local organization of decision-making . The same description could be applied to ß2−/− mice . However , in mutants , the percentage of transitions from periphery to center ( PA → CA ) was enhanced , while the “stops in the center” transitions ( CA → CI ) were reduced ( Figure 3B ) . Stationarity has been tested by comparing transition probabilities obtained during the first and the second 15 minutes of the experiment . We observed ( i ) a slight modification of ( PI → PA ) probability of transition ( it decreases from 97 . 7% to 95 . 5% , and from 98 . 0% to 96 . 0% in WT and ß2−/− respectively ) , and ( ii ) an increase of ( CA → CI ) transition with time ( from 22 . 3% to 32 . 2 and from 13 . 9% to 25 . 3 in WT and ß2−/− mice respectively ) . This last modification indicates that animals have a higher tendency to stop at the center in the second part of the experiment . This increase is similar in WT and in ß2−/− mice . Distributions of residence times were also modified in ß2−/− mice ( Figure 3C ) . Comparison of the mean of residence times in individual using the Wilcoxon test indicated that PI , PAc and PAp residence times were significantly modified in ß2−/− mice . PI average duration was reduced 13% ( Δ mean = 0 . 58 sec , p = 0 . 028 ) , while PAp and PAc average duration were increased 15 . 2 and 35 . 3% ( Δ mean = 0 . 72 , p = 0 . 0017 and 1 . 66 sec p = 1 . 7e-6 ) , respectively . Mean of CI or CA states were not statistically modified , despite an apparent difference in the distribution of CI ( not shown ) . In the state sequence , CA is preceded either by PAp , PAc or CI . In WT , there was no significant difference between time distributions of CA , depending on the preceding state ( Wilcoxon test ) . In contrast , CA resident time was increased after a CI when compared with PI preceding a PAp or a PAc ( mean = 3 . 09 against 2 . 7 and 2 . 8 sec , Wilcoxon test , p<0 . 001 in both cases ) . Similar dependencies on preceding state were observed for PI state duration . Mean duration varied significantly ( mean = 4 . 01 , 5 . 16 and 4 . 11 sec , Wilcoxon test , p<0 . 001 in pair comparison ) after CI , PAc or PAp , respectively ( mean = 4 . 01 , 5 . 16 and 4 . 11 sec , Wilcoxon test , p<0 . 001 in all pair comparisons ) . Similar properties were observed in ß2−/− mice ( mean = 3 . 08 , 4 . 32 and 4 . 08 sec , Wilcoxon test , p<0 . 001 in all pair comparisons ) . Deletion of the ß2-subunit gene affected both the residence time distribution and the transition matrix . To identify more specifically the locus of the behavioral sequence where the mutation effect takes place , we used a modeling strategy ( see Methods ) . We first checked the validity of the simulation ( see also Text S1 and Figure S1 and Figure S2 ) and that the numbers of occurrences of each of the five states in 30 min experiment agreed well in both WT and ß2−/− mice with numbers obtained with simulated data when the respective matrix of transition and residence times were used . Accordingly , the total traveled distance being almost linearly ( Figure 2C ) related to the total time spent in each of the five states , it was also well-reproduced using simulation ( Figure 4A ) . We also tested the impact of non-stationarity and resident time sequence dependency ( see also Text S1 and Figure S1 ) on the simulation . To further dissect the respective contribution of the transition matrix and of the residence time distributions , we modeled data based on: ( i ) transition matrix of WT and residence time distribution of WT ( labeled WT/WT ) , ( ii ) transition matrix of ß2−/− and residence time distribution of WT ( ß2/WT ) , ( iii ) transition matrix of WT and residence time distribution of ß2 ( WT/ß2 ) , and ( iv ) transition matrix of ß2−/− and residence time of ß2−/− ( ß2/ß2 ) , and we compared the time spent in PI and in PAc ( Figure 4B ) for the various model configurations . Convolving matrix and residence time distribution demonstrated that none of them fully explained modifications of the time spent in a given state and consequently the “hyperactivity profile” . Transition probabilities and residence time distribution explained individually no more than 56% of the total difference observed between WT and ß2−/− , while their sum effect explained 95 and 92% of the total mean difference observed between WT and ß2−/− . In terms of quantification this suggested that both matrices and distributions of residence time should be used . A final question was whether a single modification of a WT sequence property could reproduce most of the ß2−/− phenotype . The observed behavioral changes between WT and ß2−/− are open to a variety of interpretations . One of them is that ß2−/− specifically reduce some stops . The main advantage of such hypothesis is that modification of only one element ( decreased number of stop ) accounts for matrix and residence time difference between WT and ß2−/− mice . A simple simulation ( see Methods , “stop reduction” model ) revealed that removing 30% of stops in WT sequences reproduced well the number of occurrences of each of the five states ( Figure 5A ) , matrices ( Figure 5B ) , and residence time distributions ( Figure 5C ) . More precisely , PI was not changed , which means that the model does not explain the decrease observed in ß2−/− mice . However , Pap and Pac increased to a level compatible with resident time observed in ß2−/− mice ( Δ mean = 0 . 27 sec , Wilcoxon test , p = 0 . 09 and Δ mean = 0 . 43 sec , Wilcoxon test , p = 0 . 49 for Pap and Pac respectively ) . Such modeling identified the “stop” as an element that could explain differences between WT and ß2−/ . We then focused our analysis on this particular moment . Finite-state systems deriving from the discrete analysis of a continuous movement necessarily coarsen the fine structure of that movement . What has been , so far , identified as inactivity in this paper , is a mode of motion close to a complete stop of the animal . During this period of inactivity the mouse can however make a variety of movements . The animal can progress forward slowly ( with a small but constant speed ) , freeze , perform a number of action patterns ( i . e . , grooming , rearing , scratching , etc ) , or orienting movements ( head scanning , sniffing , etc ) . In order to be able to differentiate some of these patterns , we have simultaneously recorded the position of the animal and digitized video images ( 25 frames/second ) . These images have been used as the input for fine off-line movement analysis ( Figure 6A ) . Visual analysis of video images allowed us to distinguish periods with rearing and head scanning movements , from periods with only reorientation or no change in orientation . Five classes of inactivity periods were have been distinguished . They corresponded to rearing , scanning , grooming , border rearing and sniffing ( see Methods ) . Stops at the periphery of the open-field were differently distributed in WT ( n = 14 ) and β2−/− ( n = 11 ) mice ( Figure 6B ) . The numbers of rearing , wall rearing , and sniffing were not affected and were similar in both strains ( Δ = 3 . 18 , Wilcoxon test , p = 0 . 32; Δ = 0 . 4 , Wilcoxon test , p = 0 . 80; Δ = 6 . 18 , Wilcoxon test , p = 0 . 12 , respectively ) ) . Grooming patterns were increased ( Δ = 4 . 1 , Wilcoxon test , p = 0 . 003 ) , whereas scanning was decreased ( Δ = 6 . 9 , Wilcoxon test , p = 0 . 0008 ) in mutant mice . Scanning behavior being related to the “exploration” of , or the information update about , the environment , differences observed in scanning could therefore have a consequence on the sequence of behaviors . New information obtained by the splitting of PI into five subtypes identified by the dominant behavioral acts , i . e . rearing , scanning , etc . , can challenge the description of the sequences in two ways . First , the knowledge of the animal acts during a PI state can modify the probabilities of consecutive states without modifying the first-order Markov description . Second , new information about PI can modify not only the conditional probabilities but also the order of the Markov description , thus requiring a more complex description of the process . The conditional probability of transition from PA to CA was modified by the knowledge of the behavioral act performed during stops preceding PA ( Figure 6B , left ( top ) , ANOVA , F ( 6 , 91 ) = 13 . 4 , p = 8e-11 ) . More specifically , P ( CA|PA ) = P ( CA|PI-PA ) , when no further indication is given on PI , but the probability of transition was greatly enhanced when the animal performed scanning . That is , P ( CA|PA ) <P ( CA|PIsc-PA ) if PIsc was a scanning behavior ( Δ = 0 . 36 , test p = 1 . 5 e-08 ) . These results showed that after scanning an animal tended to engage more frequently in a transition to the center of the arena than after a stop paired with a different activity . Probability to stop at the center of the arena was however not modified by the activity of mice during a PI ( Figure 6B , left ( bottom ) , ANOVA , F ( 7 , 104 ) = 0 . 91 , p = 0 . 49 ) . In ß2−/− mice , the modification of probability after scanning disappeared , that is , the first order model was not modified by knowledge of the behavioral act occurring during a PI ( Figure 6B , right ) . Providing new information about the PI state modified the Markov order of the description . We therefore switched to Variable Length Markov Chain modeling ( see Methods ) . If we consider two main populations of stops , i . e . scanning and no-scanning , a tree representation of the influence of the past behavior , i . e “the context” , on a given decision can be built . For this purpose , the sequence of symbols was fitted using a Variable Length Markov Chain model ( VLMC , see Methods ) . Animal trajectories were described using six symbols CI , CA , PAp , PAc , PInsc and PIsc , the two last states coding for stop at the periphery without or with scanning , respectively . Sequences from different animals were concatenated for VLMC analysis . The WT mice context tree ( Figure 7A , left ) showed seven contexts . Five of them were first order ( from top to bottom , CI , CA , Pac , PInsc and PIsc , Figure 7A ) , indicating that the next symbol ( X ) depends uniquely on the present state . More interestingly , two contexts with second order also appeared . The first corresponded to the previous demonstration that after “scanning” an animal tended to engage more frequently in a transition to the center of the arena . The second indicated that , in contrast , when mice did not perform scanning , they preferentially made a stop in the periphery . This is schematized ( Figure 7A , right ) by a “PI choice point” , where the movements that follow depend on what activity the mouse had performed during the previous PI . The context tree of β2−/− mice was made of eight contexts , four of them ( CI , Pap , PInSC , PIsc ) being of first order . The architecture of the tree was clearly modified when compared to WT . Strikingly , dependence between movements during PI and “transition to center” completely disappeared . In contrast , the tree highlighted different chains in the ß2−/− sequence of behavior , with chains of second or third order that organized movements and relations between PAc and CA ( Figure 7B ) .
In this paper we have investigated the processes underlying ß2−/− mouse hyperactivity in an open field . These mice exhibit an increase in the total distance traveled in the open field by about 40% when compared to WT . Consistent with this hyperactive phenotype , ß2−/− mice spent more time in fast , and less time in slow , movements . To analyze mouse trajectories we developed a specific approach based on a dissection of mouse behavior in the open field as a sequence of motor activities organized in patterns . We have shown evidence for two main modifications of the behavior in ß2−/− mice: ( i ) quantitatively , mutant mice show a reduced number of stops and modification of specific transition probabilities , and ( ii ) structurally , the organization of the sequence of behavior was different between strains . Streams of complex acts or movements exhibit some regularity that is the basis of the subdivision of behaviors into units , or species-specific movements . In rodents , a variety of complex sequences of action have been identified [20] . In our analysis we focused on two classifications , active versus inactive , and central versus peripheral movement . Although simple , this classification captures two essential and ethologically meaningful properties of the displacement . The first is the alternation between progressions and stops , observed in a number of locomotor behaviors , and associated with prey search , vigilance or energy saving [21]–[23] . The second concerns the spatial distribution of movement . Traveling close to the wall is an important feature of the mice , and it has been suggested that the wall confers security while the center is anxiogenic . However , exploratory behaviors also drive the mouse to explore all the open space . A more precise definition of the different movements can be performed [15] , [24] , but our coarse-grained decomposition allowed us to focus on sequence properties , and to obtain sufficient stationary data in 30 min experiments , for a robust statistical description of simple spontaneous decision making ( engage in the center of the arena , stop… ) . Analysis of behavior in terms of sequences and Markov processes has been already applied to different species [25] . Markov analysis assumes that the underlying process that generates a sequence is homogeneous in time all along the sequence . The time range over which an event influences the future ones is supposed to be constant ( i . e independent of the event and the sequence preceding it ) . For this reason , fixed length Markov chain analysis is a poor detector of sequence rules that operate only after a particular portion of the sequence . By contrast , VLMC allows identification of particular sequences or contexts , such as those identified after scanning an environment . Modification of this homogeneity in sequences is often seen as an indicator of higher organization such as “hierarchical” or “grammatical” properties [26] , [27] or reflects specific ‘decisions’ [26] . The methodology applied in this paper is not intended to be a blind modeling but rather a way of testing hypotheses , giving or not significance to ‘a priori’ choices and categories . It offers the possibility of including ethological knowledge and previously established categories . It would then also be relevant and efficient also in more naturalistic and complex settings . The VLMC framework can be generalized so as to investigate whether the grouping of categories in classes is relevant . It thus proves to be useful to improve the parsimony of the description [28] . Hyperactivity in an open field can take different forms , including faster locomotion , longer periods of travel , fewer pauses , shorter pauses , etc . The question is then whether the reduction of the number of stops is sufficient to explain the hyperactive profile . Our experiments demonstrate that locomotion is not faster in ß2−/− mice , and that the difference lies in the patterns and organization of behaviors . Furthermore , a simulation approach suggests that hyperactivity cannot be explained only by changes in the matrix , or only by changes in the duration of the various states , but by their joint effect . Hyperactivity would then emerge from alterations of many different underlying processes . However , we here propose that in ß2−/− mice hyperactivity is mainly due to the “lack of stops” . Most characteristics of the sequences of ß2−/− mice can be explained by the fact that these mice do not observe certain “stops” and that after a stop they organize their behavior differently . The significance of such a modification and the underlying changes it reflects is , however , not trivial . Open-field behavior , also called exploratory behavior or locomotor behavior in a novel environment has been initially used as an indicator of anxiety/emotionality [10] , [11] . It is also used to study exploration and how animal react to novelty , an approach with known limitations [10] , [13] , the most important difficulty being that the various open-field measurements do not represent a single dimension of behavior ( i . e , emotionality or exploration ) . This limitation reinforces the interest of using sequence analysis , which does not make any assumptions about any underlying process , but focuses on the organization of behavior ( see also [29] ) . Most important features of an animal's displacement organization can be summarized as follows ( Figure 8 ) : At the periphery , after a “stop” , the probability that WT mice engage movement in the center of the arena is 36% . This probability is ( i ) increased by “scanning” ( up to 61% ) and ( ii ) decreased by a recent excursion to the center ( down to 24% ) . In ß2−/− mice this probability is different in baseline ( 48% ) , the increase caused by scanning disappears and the decrease by recent incursion is similar . These results point to information gathering as a key element underlying differences between WT and ß2 in the organization of sequence of behavior in an open field . The ability to adapt to an unfamiliar or uncertain environment is fundamental , and an essential point in adaptation would be that animals actively look for a modification in the environment . Displacement of an animal in a novel environment is characterized by intermittent locomotion , scanning , and pauses that can be used to gather information about environment but also to reduce unwanted detection by an organism's predators [22] . Organization of locomotor behavior in an open environment is compatible with optimization theory insofar as it minimizes risk while maximizing gain , i . e . collect information about environment [30] . Fear and anxiety tend to reduce center movement , while exploratory motivation tends to increase these movements [24] . Accordingly , increased probability of center engagement after scanning may be viewed as caused by a reduction of anxiety ( Figure 8 ) . Yet , WT and ß2−/− mice have similar levels of anxiety [4] , [31] , furthermore the parallel evolution of CA → CI probability of transition suggest that reduction of anxiety with time is similar in both strain . The observation that the structure of the displacement is modified in ß2−/− mice and that this modification targets “scanning” as a key feature in the organization of behavior suggests instead a modification of information gathering and of the risk/gain optimization . The notion that exploratory behaviors in novel environments may serve to optimize safety and that this behavior is modified in ß2−/− mice also parallels previous observations suggesting that WT mice react to novelty by increasing exploratory activity , whereas ß2−/− mice do not adapt their behavior to a change in the environment [4] . It has been proposed that the alteration of behavioral adaptation in ß2−/− mice , coupled with unimpaired memory and anxiety , may model cognitive impairment observed in human disorders [4] such as attention-deficit hyperactivity disorder ( ADHD ) [32] , or even in autism [5] . This proposition relies upon the idea that behavioral flexibility is controlled by an adequate hierarchization of motivations , a process known to mobilize prefrontal and cingulate cortex . ADHD symptoms such as inattention lack of inhibitory control , and hyperactivity and prefrontal involvement indeed resemble ß2−/− behavioral deficits , and fit well with nAChR localization and function . Yet , the possible contribution of prefrontal cortex and higher-level top-down processes in open-field behaviors is at this stage not clear . More complex environments and tasks , together with relevant methods of analyses , are needed to explore this problem . Further experiments are also needed to clearly identify the brain loci and the nicotinic receptor subunits that are involved in the modification of the behavioral patterns observed in ß2−/− mice . This fine-tuned analysis of the way wild-type and mutant animals organize their spontaneous activity may ultimately help to understand the contribution of nAChRs to higher brain functions in humans , and the abnormalities that accompany many neuro-pathologies .
Exploratory activity was recorded in a 1-m diameter circular open-field . Experiments were performed out of the sight of the experimenter and a video camera , connected to a Videotrack system ( View-point , Lyon , France ) , recorded the trajectory of the mouse for 30 minutes . To characterize stopping behavior ethologically , home-made softwares ( Labview , National instrument ) were used to acquire film with a higher resolution . Initially introduced in a purely mathematical context , symbolic dynamics has also been developed as an efficient tool for data analysis [33] . It provides a framework to investigate generic features of a dynamical system from the knowledge of experimental trajectories , in particular when only short series are available , when individual variability is important , or when only a few features within the recording are relevant . The core idea is to encode continuous-valued trajectories into behaviorally relevant symbol sequences associated with a finite partition of the state space . Velocity and position of the mice were used to define a partition in four states ( or symbols ) , by combining two binary ones ( see below ) : When combined , these symbols give four codewords or states {PA , PI , CA , CI} that correspond to Activity or Inactivity in the Periphery or in the Center of the arena . Animal trajectories in the open-field are then represented by a sequence of codewords ( Figure 1C ) . The choice of a specific threshold value to partition symbols and the range of validity of these values have been discussed and analyzed in a previous paper ( see Supporting Information [6] ) . The 2-D paths were smoothed using triangular filter . The instantaneous velocity can be then meaningfully computed from these smoothed data , simply implementing its definition ( first time-derivative of the position ) . Instantaneous velocity range was partitioned in two sub-ranges delineated by the threshold θ1 . A second threshold θ2 has to be involved in order to faithfully assess activity , according to the following rule:allowing to encode the continuous trajectory into a binary sequence φv ( t ) . In other words , it means that crossing the low threshold θ1 can be considered as the starting point of a significant active phase if and only if the velocity reaches the high threshold θ2 . This high threshold determines qualitatively the active type of the period whereas the low threshold determines quantitatively its duration . This dual criterion avoids spurious alternation of active and inactive phases of arbitrary small duration . Indeed , since the acceleration of the mouse is bounded above by some value amax , the duration of an active phase is at least ( θ2-θ1 ) /amax hence the choice of the thresholds implicitly fixed a lowest bound on the time scales . In fact , a lowest bound on the time scale was also prescribed explicitly: an additional temporal smoothing achieving a stronger masking of fast velocity fluctuations is performed by fixing a minimal duration above or below the low threshold to record it as an actual crossing . The two-threshold criterion masks the presence of weak peaks in the velocity that do not overwhelm significantly θ1 ( even if they last long ) while the explicit constraint on duration masks the narrow peaks ( fast fluctuations ) even if they reach high velocity values . The combination of these two criteria moreover ensures that the resulting binary sequence is not very sensitive to the precise value of θ1 ( this feature has also been checked directly ) . The area of the arena was divided in two regions , with a central zone C ( Center ) with Rc<1 and an annulus P ( periphery ) . Then , depending on the continuous radial position R ( t ) = ( x2+y2 ) 1/2 , defined in such a way that it ranges from 0 to 1 depending on whether the mouse was close to the border of the arena ( R = 1 ) or at its center ( R = 0 ) , the trajectory of the mouse is transformed into binary sequence φp ( t ) by: In this study Rc = 0 . 65 . In order to be able to differentiate patterns of inactivity , video of the animal displacement was recorded ( 25 frames/second ) and used to detect the position of the animal . To classify the stops without bias , only parts of the movie considered as PI in the behavioral sequence were watched without looking either at the duration of the stops , or at the following sequence . We used five classes of behavior for this classification , rearing , grooming , border rearing , sniffing and scanning [20] . Such an ethological classification has been chosen for its clarity as regarded the aims of the different behaviors . Grooming is defined by a well-characterized sequence beginning with movements of paw cleaning and proceeding through face washing and body cleaning . Rearing and border rearing were easy to distinguish , the animal raises upon its back paw . Difference in between rearing and border rearing is whether front paw touch the border of the open-field or not . Sniffing is defined by an activity in which the mouse sniffs the ground , this behavior is usually used to identify object or food or to make spatial landmark . Scanning contains any information gathering about the environment , beginning with rearing but the animal then engages large head movement that can be accompanied by sniffing . Henceforth , we shall call “symbol” each of the 4 codewords PA , PI , CA , CI since the binary symbols will never be considered in isolation in what follows . One way to analyze a sequence consists in analyzing the probability of transition from one state to another . From the initial time series written with an alphabet of x symbol , a x*x matrix T = ( tij ) can be calculated , where tij is the number of times a given symbol i is followed by another symbol j in the sequence . T is called a transition frequency matrix . A conditional transition matrix can be obtained by dividing each row of the transition frequency matrix by its sum . Conditional probabilities for each state are then estimated by unbiased estimator p ( A|B ) = n ( BA ) /n ( B ) where ( n ( BA ) designates the number of 2 symbol sub-sequences where B is followed by A . Transition frequency matrices and conditional transition matrices are a concise way of expressing the statistical relationship between consecutive states . They give preliminary clues to the organization of the sequence of states . This is generally summarized in a flow diagram , giving a simple graphical representation of these matrices . Nodes in the diagram represent states , while arrows of variable thickness represent the frequencies with which the different transitions occur . This representation provides a suitable overview of the organization of the sequence of behaviors ( see Figure 3 ) . The matrix of transition describes the statistics of transitions from one state to the other but it does not provide any information about the dynamic nature of the relationship between successive states . Obtaining information about the dynamics in short and long terms from the sole knowledge of the transition matrix is possible only if the dynamics is Markovian: A process is a first-order Markov chain if the transition probability from state A to the next state B depends only on the present state A and not on the previous ones . A first-order Markov model is then a mathematical model fully prescribed by the transition matrix that describes , in probabilistic terms , the dynamic behavior of the system , namely the probability of transitions over any duration between any two states . In such a model , the present state contains all the information that could influence the choice of the next state , that is captured in the transition matrix . A classical way to demonstrate that a process is Markovian is to show that the sequence cannot be described by a zero order process , i . e . that P ( B|A ) ≠ P ( B ) and that P ( C|B ) = P ( C|AB ) , but see [25] for a more detailed review of all these methods . The residence times , defined as the time spent in a given state , were studied separately . We described the dynamics of transition between states using an alternate renewal process . That is the sequence is described by the convolution of a Markov chain describing the transitions between the states associating a unit time step to each transition , with the above residence-time distributions , describing the actual duration of each step . Thus , there is no repetition of states in the sequence and the transition matrix has vanishing diagonal elements The most interesting part of the Markov formalism is that the knowledge about the transition probability , i . e . the elementary properties of the system , is sufficient to describe the whole dynamics of the system , either in the short or long term . In practice , this means that as soon as a first-order Markov process has been demonstrated , modifications induced by drugs , genetic mutation or other manipulation of the system can be localized in the transition probabilities and/or in the time distribution of state duration ( provided the investigated perturbation does not affect the first-order nature of the dynamics ) and the same modeling strategy can be used . Modeling procedure is as follows . We used ( i ) the conditional probabilities from a given state to specify the next one , and ( ii ) the residence time distributions to determine durations of the successive states . This whole procedure is reiterated until the total duration reaches half an hour of experiment . These synthetic data can then be compared with those obtained experimentally . In a second time , specific modification of transition probabilities or residence time distributions are used to access impact of such a modification . A specific model , consisting in “stop reduction” has been particularly used . In this model , sequences of symbols are generated using WT matrices and distribution . In a second step a fixed percentage of stops ( 35% of both PI and CI ) are removed in such a way that PA-PI-PA becomes PA-PA , that is a unique PA event but with a longer duration ( and similarly for CA-CI-CA ) . The total length of the sequences is adjusted in a way that it represents a half-an-hour experiment . When the dynamics is not accounted for by a first-order Markov chain , but displays larger dependence on the past states , “variable length Markov chains” ( VLMC ) provide an efficient modeling [19] . In this class of models , dynamics is still prescribed by the expression of conditional probabilities of the future states . But now , each history from t = −∞ up to time t is truncated into finite sequence from t-s to t , with s≥0 , having actually an influence onto the states at time t+1 . For all B , P ( B at t+1 | past up to t ) = P ( B at t+1 | C ( past , t ) ) . The length of the truncated sequence C ( past , t ) , called a context , depends on the history instead of being uniformly equal to the length of the longest one . The gain in reducing the dimension of the parameter space is obvious when the dynamic memory is heterogeneous ( context-dependent ) . A VLMC is thus characterized by: ( i ) a set of finite-length context , and ( ii ) a family of transition probabilities associated to each context . The context defines the finite portion of the past that is relevant to predict the next symbol ( whatever it is ) . Given a context , its associated transition probabilities define the distribution of occurrence of the next symbol . VLMC analyses were performed on concatenated chains obtained from different animals of the same group . The R-package VLMC was used to fit data . Fittings were performed in two steps . First a large Markov chain is generated containing the context states of the time series . In our analysis only nodes that appear n = 5 times per animals ( that is 70 for 14 WT and 55 for 11 β2−/− ) were taken into account to generate the initial tree . The obtained results are almost insensitive to the value of this parameter n . In the second step , many states of the Markov chains were collapsed by pruning the corresponding context tree . The pruning requires definition of a cutoff value . A large cutoff yields a smaller estimated context tree . In our analysis cutoff value corresponding to 1‰ was used in order to extract strong and significant contexts . All data were analyzed using R , a language and environment for statistical computing . Data are plotted as mean±95% confidence intervals . Boxplot is also used when information about distribution is important ( see Figure 2A and 2B , for example ) . Boxplot summarizes data using the smallest observation , lower quartile ( base of rectangle ) , median ( line in rectangle ) , upper quartile ( summit of rectangle ) , and largest observation . Data points considered outliers are marked by isolated points ( circle ) . Total number ( n ) of observations in each group and statistics used are indicated in figure captions . Classically comparisons between two means are performed using two-sample t . test . When there is doubt about the normality of the data distribution , non-parametric Wilcoxon rank-sum test is preferred . For variable Markov chain model fitting , VLMC package is used .
|
Understanding mechanisms underlying complex behaviors and the abnormalities that accompany most neuropathologies is a current challenge in biomedical research . A number of approaches is primarily based on the identification of genes and their associated molecular pathways implicated in complex motor or cognitive pathologies . However , optimal use of the large body of genetic , molecular , electro-physiological , and imaging data is hampered by the practical and theoretical limitations of currently available behavioral analysis methods . Complex behaviors consist of a finite number of actions combined in a variety of spatial and temporal patterns . In this paper we develop a sequential analysis of mouse displacement in an open-field paradigm and demonstrate that a description based on a Markov model can be used to describe quantitatively patterns of behaviors and to detect changes in the way that animals organize their displacement , especially in mice lacking nicotinic acetylcholine receptor subunits . This paper would be of broad interest not only to those concerned with this particular mice model but also generally to those interested in modeling complex behavior traits in mice .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"neuroscience/behavioral",
"neuroscience",
"neuroscience",
"neuroscience/animal",
"cognition"
] |
2008
|
Behavioral Sequence Analysis Reveals a Novel Role for
ß2* Nicotinic Receptors in Exploration
|
Bacterial operons synchronize the expression of multiple genes by placing them under the control of a shared promoter . It was previously shown that polycistronic transcripts can undergo differential RNA decay , leaving some genes within the polycistron more stable than others , but the extent of regulation by differential mRNA decay or its evolutionary conservation remains unknown . Here , we find that a substantial fraction of E . coli genes display non-uniform mRNA stoichiometries despite being coded from the same operon . We further show that these altered operon stoichiometries are shaped post-transcriptionally by differential mRNA decay , which is regulated by RNA structures that protect specific regions in the transcript from degradation . These protective RNA structures are generally coded within the protein-coding regions of the regulated genes and are frequently evolutionarily conserved . Furthermore , we provide evidence that differences in ribosome densities across polycistronic transcript segments , together with the conserved structural RNA elements , play a major role in the differential decay process . Our results highlight a major role for differential mRNA decay in shaping bacterial transcriptomes .
One of the defining features of bacterial gene expression is the co-transcription of multiple genes within polycistronic mRNAs [1] . Co-transcribed genes often encode proteins that participate in the same biological process or interact in a protein complex , allowing these genes to share the rate limiting steps of transcription . Thus , polycistronic transcription can synchronize the expression of functionally related genes , streamlining gene regulation in the compact genomes of bacteria . A major limitation of polycistronic transcription is that it theoretically generates equal amounts of mRNA for all genes sharing an operon . This can be disadvantageous in the common scenario where one of the proteins encoded in the operon is needed in higher amounts than the others . In many such cases this inconsistency is resolved by differential translation efficiencies of individual genes encoded on the same transcript [2 , 3] , or by internal transcription start sites ( iTSSs ) in the middle of the operon [4] . However , while differential translation efficiencies and iTSSs have been identified in many transcriptional units [2 , 3 , 5 , 6] , differential operon expression patterns cannot always be explained by such mechanisms . More than three decades ago it was shown that mRNA degradation can lead to differential mRNA stoichiometries for genes encoded on a single polycistron [7 , 8] . This regulation by differential mRNA decay was demonstrated for the E . coli malEFG operon encoding the maltose ABC-transporter complex , where MalF and MalG form the transmembrane channel and MalE acts as the substrate-binding protein , which is needed in higher quantities [8] . It was shown that the 3’ portion of the polycistronic mRNA ( including malFG ) is rapidly degraded by RNases whereas the 5’ portion ( malE ) is stabilized , leading to accumulation of the malE mRNA and to a substantial relative increase of the MalE protein as compared to MalF and MalG [8] . While regulation via differential operon decay was reported in multiple bacterial operons over the last few decades [9–17] , the extent of this regulatory mechanism or its evolutionary conservation in bacteria remains unknown . In this study , we combine a set of multi-layered high-resolution RNA-seq approaches to extensively map and characterize differentially decaying operons in E . coli . We find that regulation by differential decay is widespread and reshapes the stoichiometries of ~12% of polycistronic mRNAs in this model organism . This process is highly conserved also in the Salmonella and Enterobacter transcriptomes , and depends on conserved RNase-resistant RNA structures that guide the RNA decay dynamics . Furthermore , we find that differential decay is often dependent on translation and that stabilized operon segments are characterized by relatively higher ribosome densities than non-stabilized segments , suggesting that variation in translation efficiency guides endonuclease cleavage to ribosome depleted sections . Taken together , our data support differential operon decay as a common and frequently conserved mode of regulation in bacteria .
To examine the possible extent of differential RNA degradation in shaping operon stoichiometries , we first sought to limit our analysis to a set of well-defined E . coli operons annotated in the EcoCyc database [18] . We discarded from the set operons that were expressed at low levels or that are likely regulated by iTSSs or by internal partial transcription termination under the conditions tested in this study ( Methods ) . For this , we comprehensively mapped the TSSs in exponentially growing E . coli bacteria using a previously established approach [18 , 19] ( S1 Table ) , and excluded cases suspected as regulated by iTSSs ( Methods ) . We further excluded cases where two consecutive genes within the operon were separated by an intrinsic terminator motif ( a stem-loop structure followed by a uridine tract ) or in which substantial Rho-dependent termination was previously described [20] ( Methods ) . The final set included 390 significantly expressed multi-gene operons overall encompassing 1292 genes , representing 31% of the genes in the E . coli BW25113 genome ( S2 Table ) . We then treated exponentially growing E . coli cultures with rifampicin , a transcription initiation inhibitor [21] , and tracked the RNA decay process by sequencing mRNAs collected at 1-minute intervals following transcription inhibition [22 , 23] ( S1 Table; Methods ) . These data were used to calculate the half-life of individual genes ( average half-life = 1 . 5 +/- 0 . 9 minutes ) , with the results overall agreeing well with recent RNA half-life calculations for E . coli genes [22 , 23] ( S1 Fig; Pearson R = 0 . 75 , p < 10−15; Methods ) . However , we detected 47 operons ( 12% of the 390 analyzed operons ) , encompassing 177 genes , in which two consecutive genes displayed correlated differential decay rates and differential steady-state mRNA levels ( Fig 1A–1D; S3 Table; Methods ) . Re-analyzing RNA decay data from recent studies corroborated our results [22] ( S3 Table ) . It is well established mathematically and experimentally that the ratio in steady-state mRNA levels of two equally transcribed genes should equal to the ratio in their half-lives ( for example , if the half-life of gene A is 2 minutes and the half-life of gene B is 1 minute , then at steady-state gene A will be 2-fold overexpressed as compared to gene B ) [24] . Indeed , for the majority of operonic gene-pairs in which we found differential decay rates , the estimated differences in half-life closely matched the observed differences in steady-state mRNA abundance ( Fig 1D; S3 Table ) . In about one quarter of the cases we found that the differences in decay rates did not fully match the differences observed in the mRNA level , indicating either measurement error or additional regulation by iTSSs or leaky termination in these operons ( S3 Table ) . Among the decay-regulated operons we identified was the previously described maltose ABC transport operon , malEFG [8] ( Fig 1A ) . We measured the half-lives of the malFG and the malE segments to be 1 and 3 . 5 minutes , respectively , consistent with the relative stabilities measured for these genes in previous reports and validating our measurements . In addition to the malEFG operon we identified four other ABC transport systems , including those transporting polyamines , arginine , histidine and methionine ( Fig 1B and 1C; S3 Table ) . In all these operons , we found that the stabilized gene was invariably the substrate-binding subunit , which is typically needed in higher quantities than the channel-forming units . This effect was independent of the gene’s position within the operon . For example , while the stabilized maltose substrate-binding subunit gene , malE , is the first gene in the operon ( Fig 1A ) , the substrate-binding subunits of the polyamine ( potD ) and arginine ( artI ) transporters occur at the end or the middle of their operons , respectively ( Fig 1B and 1C ) . Stabilization of the mRNA of the substrate-binding subunit results in higher steady-state mRNA levels of the respective gene , suggesting that differential decay is a common mechanism for tuning differential stoichiometries in ABC transport systems . Another transport system subject to decay-based regulation was the tatABC operon , which encodes the evolutionarily conserved twin-arginine protein export system [25] ( Fig 1D ) . In this system the TatA protein forms homo-polymeric ring structures that contain on average 25 TatA copies for each TatBC complex [25 , 26] , and we indeed found the mRNA segment encoding the TatA subunit to be highly stabilized as compared to that encoding the TatBC subunits ( S3 Fig ) . We also observed differential decay in the rpoZ-spoT containing operon , such that the segment encoding rpoZ is stabilized compared with that containing spoT . The rpoZ gene encodes the RNAP omega subunit , which binds the regulatory alarmone ppGpp [27] , and SpoT is one of the major enzymes involved in the synthesis and degradation of this alarmone [28] ( S3 Table ) . As the enzymatic activity of a single SpoT rapidly converts multiple ppGpp molecules and hence can affect multiple rpoZ gene products , it is likely that RpoZ protein expression levels need to be higher than that of SpoT . Conceptually similar , we also identified differential decay in operons encoding 4 different two-component systems , where in each operon the transcription factor component was consistently more stable than the histidine-kinase enzyme [29] . Finally , we identified additional differentially decaying operons involved in cellular signaling and regulation , protein translocation across the membrane , antibiotic resistance and various metabolic processes , suggesting that differential decay plays an important role in post transcriptional regulation of many physiological processes in E . coli ( S3 Table ) . To examine whether regulation by differential decay is evolutionarily conserved we performed an identical rifampicin-based decay assay using the human pathogenic bacterium Enterobacter aerogenes grown under the exact same conditions as E . coli ( S1 Table; Methods ) . We found that 71% ( 20/28 ) of decay-regulated orthologous operons expressed in E . aerogenes under the tested conditions show conserved differential decay patterns ( S2 Fig; S4 Table ) . In addition , we compared the steady-state mRNA levels of E . coli decay-regulated operons with that of their orthologous operons annotated in Salmonella typhimurium , grown under the same conditions . This analysis showed that 68% ( 26/38 ) of the decay-regulated E . coli operons shared between the two species present similar patterns of sub-operon differential expression in S . typhimurium ( S3 Fig; S1 and S5 Tables; Methods ) . The observed conservation of differential decay patterns between different bacteria highlights the functional importance of this phenomenon in bacterial gene regulation . Our analyses discover many new cases where the steady-state levels of individual genes within polycistronic mRNAs are controlled via selective stabilization ( differential operonic decay ) ( Fig 1; S3 Table ) . However , why some transcript regions are protected from digestion whereas others are rapidly degraded was unclear . In the well-studied maltose operon , it was shown that the malE gene is stabilized due to the presence of a 3’ protective RNA structure that resides in the intergenic region between malE and malF . This structure exerts its protective effect on malE following an initial endonucleolytic cleavage event that generates a functional malE fragment still physically attached to the protective element . This structure can then resists the 3’-5’ exonuclease processive activity typically performed in Gram negative bacteria by the Polynucleotide Phosphorylase ( PNPase ) and RNase II enzymes [30 , 31] , thus stabilizing the malE gene , which is the 5’-most gene in the operon [8 , 32] ( Fig 2A ) . Similar stabilization via 3’ RNA structures was also described in a few additional operons [9 , 10] . The set of differentially decaying operons we found included 34 gene-pairs ( 70% ) in which the 5’-most gene is stabilized as compared with the downstream region , akin to the case of malEFG ( Fig 1A; S3 Table ) . In 76% of these ( 26/34 ) , the 3’ end of the stabilized mRNA portion ( as determined by the term-seq method [33] ) occurred immediately downstream to an energetically stable RNA-hairpin ( Fig 2B–2D; S6 Table; Methods ) . These RNA structures were significantly more stable than those occurring randomly across the genome ( S4 Fig; p < 10−14 , Wilcoxon ) and were not followed by the uridine tract required for Rho-independent termination [34] ( S6 Table ) . In addition , we could not detect significant Rho-dependent termination at these sites using RNA-seq data from bicyclomycin-treated bacteria [20] , supporting that these RNA structures have a role in 3’-5’ exonuclease resistance rather than transcription termination . In the majority of cases ( 17/26 ) , the structural RNA elements presumed to protect the upstream gene from degradation were embedded within the protein-coding sequence of the downstream gene , positioned on average 107 +/- 56 nt into the coding sequence ( Fig 2B–2D; S6 Table ) . The sequence at the ORF of the downstream gene therefore carries , in addition to the protein-coding information , also the information guiding the differential decay of the transcript and thus , its stoichiometry at steady state . Examining the homologous genes in the Salmonella and Enterobacter genomes revealed that these protective RNA structures are conserved in 75% and 91% of the differential decay instances shared between E . coli and Salmonella , or E . coli and Enterobacter , respectively ( S4 and S5 Tables ) . This correlated with an enhanced sequence conservation at the wobble codon positions that overlap protective RNA structures ( p = 0 . 0008; S5 and S6 Figs; Methods ) . Interestingly , multiple protective structures can be observed in a single operon , generating complex decay patterns: for example , in the menBCE operon , we detected consecutive protective hairpins downstream of both the menB and menC genes , in correlation with the gradual decrease in stability and mRNA abundance detected in this operon ( Fig 2D ) . In 30% of the decay-regulated operons we detected , the stabilized gene of the operon occurred in the middle of the transcript or was the most downstream one ( e . g . , Fig 1B and 1C; S3 Table ) . In these cases , protection from exoribonuclease activity cannot fully explain the observed stabilization , because no 5’-3’ exoribonucleases are known to exist in E . coli [30] . However , a similar downstream stabilization pattern was previously reported in the papBA operon of uropathogenic E . coli , encoding a transcription factor , PapB , and the major pilus protein , PapA [12] . It was found that the RNase E endoribonuclease cleaves the papBA mRNA at the intercistronic region separating the papB and papA genes , at a site located upstream of an RNA hairpin structure [36] . Consequentially , this processing event produces two transcript species: the unstable papB segment , which is rapidly degraded by 3’-5’ exonucleases , and the papA mRNA , which is protected by the 5’ RNA-structure from additional RNase E cleavage , as well as an RNase resistant terminator structure at its 3’ end [36] ( Fig 3A ) . To examine whether similar 5’-protective structures also occur in the decay-regulated operons described above , we conducted experiments with WT and a temperature-sensitive RNase E mutant , which is inactivated in the non-permissive temperature of 44°C [37 , 38] ( Methods ) . We used 5’-end RNA sequencing to compare the repertoire of exposed mono-phosphorylated mRNA 5’-ends in the WT and mutant strains following brief incubation in 44°C , and identified RNase E cleavage sites as 5’-ends that were present in the WT but were consistently and substantially depleted in the inactivated RNase E mutant across three biological replicates ( S1 Table; Methods ) . In 67% ( 10/15 ) of the operons described above , a clear RNAse E cleavage site was detected immediately upstream of the stabilized gene ( Fig 3B–3D; S7 Table ) . All cleavage sites occurred in unstructured mRNA sections and at sequences closely matching the RNase E RN|WUU cleavage motif , with the conserved +2 uridine present in all cases [38] ( Fig 3; S7 Table ) . Moreover , in 7 of these 10 operons , the cleavage site occurred closely upstream of a stable RNA structure , supporting a potential papBA-like protection mechanism in these cases ( Fig 3B–3D ) . We found that all of the detected 5’ protective structures reside within the protein-coding region of the upstream unstable gene ( Fig 3B–3D; S7 Table ) . For 80% ( 8/10 ) of these operons , an identical or closely positioned RNase E cleavage site , as well as protective structures , could be detected in S . typhimurium , based on recently published in-vivo cleavage maps , indicating a high degree of evolutionary conservation at the mechanistic level [38] ( S7 Table ) . While the RNA structures described above can insulate specific transcript regions during active degradation , the endonucleolytic cleavage events that initiate differential decay must first be directed to particular operon segments . For example , in the maltose operon , RNase E must preferentially cleave the mRNA encoding malFG , but not that of malE , such that the malE mRNA will remain physically connected to its protective RNA structure [8 , 32] ( Fig 2A ) . However , the guiding factors that direct initializing cleavage events are poorly understood , even in the well characterized example of the malEFG operon [8 , 30 , 32] . Ribosome densities were previously found to positively correlate with mRNA stability , presumably by physically restricting access to RNase cleavage sites [39–42] . Re-analyzing recently published ribosome profiling data [3] , we found that in almost all differentially decaying gene-pairs the stabilized gene was covered by substantially more ribosomes than the non-stabilized genes in the operon ( after normalization to transcript levels ) , with a median of 4 . 7-fold more ribosomes per transcript coating the stabilized mRNA segments ( Fig 4A; Methods ) . Furthermore , such differential ribosome density profiles were significantly enriched in decay-regulated operons ( Fig 4A; p < 10−8 , Wilcoxon; Methods ) , implying that ribosome densities play a common role in shaping the differential decay process , possibly by decreasing endonucleolytic cleavage rates in operon regions populated by more ribosomes . To directly examine whether differential ribosome density is involved in guiding differential decay , we analyzed recently published data in which E . coli mRNA half-lives were measured following a brief pre-exposure to kasugamycin , a translation initiation inhibitor [22] . In this experiment , kasugamycin treatment is expected to result in polycistronic transcripts where all genes are equally devoid of ribosomes , providing an approach to study the contribution of differential translation to operon decay ( Fig 4B ) . Notably , the short inhibition of translation initiation resulted in a substantial reduction in differential decay in the vast majority of the regulated operons in our set ( Fig 4C; S8 Table; Methods ) , providing evidence that differential decay within operons is often dependent on differences in translation efficiency . These results suggest that differences in ribosome densities guide the endonuclease cleavage events that initiate the differential decay process within polycistronic transcripts . Combined with the results from the above sections , we chart a general model for differential decay of polycistronic transcripts in E . coli ( Fig 5 ) .
Differential decay of polycistronic operons enables bacteria to reshape uniform transcription into differential expression . This process has been studied for the last 3 decades in multiple different species and operons , including in malEFG [8] , lacZYA [15] , focA-pflB [43] , and iscRSUA [10] in E . coli K12; papBA [12] and cfaAB [11] in pathogenic E . coli; pldB-yigL in Salmonella [13]; and pufBALMX [9] in Rhodobacter capsulatus . In the current study , we took a transcriptome-wide approach to extensively map differentially decaying operons in E . coli . We find that differential decay is a common mode of regulation in bacteria , involved in re-shaping the stoichiometry of at least 12% of the E . coli operons , in a manner conserved between related species . We note that this number may actually be higher as recent studies reported the existence of condition specific differential decay [10 , 13] . The differential decay data produced in this study were organized into an interactive online browser that is available at: http://www . weizmann . ac . il/molgen/Sorek/operon_decay/ . We show , using a combination of 3’ and 5’ RNA-termini sequencing , that protective RNA structures occur at the boundaries of stabilized operon segments in the vast majority of cases and are generally encoded within the protein-coding regions of the flanking , unstable genes . In addition , we find that the less stable genes in the operon are covered by fewer ribosomes per transcript and that translation plays a major role in shaping the differential decay process . While the relation between ribosomes and mRNA decay has been well-established for monocistronic mRNAs [39–41 , 44] , our results extend this concept to multi-gene operons with differentially decaying transcript segments . Importantly , these observations provide a potential explanation for how specific operon segments are selected for initial endonucleolytic cleavage by RNases with degenerate target motifs , a piece of the mechanism that was so far less understood ( Fig 5 ) . Interestingly , stable structures at protein-coding mRNA regions were previously suggested to reduce translation efficiency [44–46] , implying that protective structures could actually play a dual role: first , blocking translation initiation , which reduces ribosome density on the flanking gene and exposes the region to increased RNase E dependent cleavage , and second , direct protection of the stable transcript region from the decay process . Although our differential decay models provide a potential mechanism for most of the regulated operons in our dataset , additional factors , such as trans-acting ncRNAs , have been found to play a role in shaping operon decay patterns by modulating access to rate-limiting endonuclease cleavage sites [10 , 13] . Notably , such ncRNAs enable condition-specific stoichiometric regulation in operons that are otherwise degraded uniformly . Indeed , our analysis failed to detect differential decay in both the iscRSUA and the pldB-yigL operons , which were recently shown to be regulated by ncRNAs activated under conditions other than the ones employed in our study [10 , 13] . Thus , considering the large number of trans-acting ncRNAs and antisense RNAs in bacteria , the extent of differential decay-based regulation is likely even greater than our current estimates . Whereas our proposed models for differential decay likely hold for other organisms that share similar RNA decay machineries ( especially proteobacteria that rely on RNase E and 3’-to-5’ exonucleases similar to E . coli ) , many bacterial lineages harbor different RNase combinations and properties , for example 5’-3’ processive exonucleases in Firmicutes [30] . Presumably , differential decay in such organisms may rely on different principles or additional molecular signals .
Escherichia coli BW25113 , Salmonella enterica subsp . enterica serovar Typhimurium SL1344 and Enterobacter aerogenes KCTC 2190 were cultured in LB media ( 10g/L tryptone , 5g/L yeast extract 5g/L NaCl ) under aerobic conditions at 37°C with shaking to an optical density ( OD600 ) of 0 . 5 . Prior to sample collection , 1:10 ice-cold stop solution ( 90% ethanol and 10% saturated phenol ) was added and the cultures were immediately placed on ice to stop all cellular processes [22 , 47] . Bacterial pellets were collected by centrifugation ( 4000 rpm , 5 min , at 4°C ) ; flash frozen and stored in -80°C until RNA extraction . For RNA isolation , frozen pellets were thoroughly resuspended and mixed in 100μl lysozyme solution ( 3mg/ml in 10mM Tris-HCl and 1mM EDTA ) pre-warmed to 37°C and then incubated at 37°C for 1min . The cells were then lysed by immediately adding 1ml tri-reagent ( Trizol ) followed by vigorous vortexing for 10s until solution is cleared . Following an incubation period of 5min at room temp ( RT ) , 200μl chloroform was added and the sample was vortexed for another 10s until homogeneous . The sample was incubated for 2-5min at RT until visible phase separation was observed and then centrifuged at 12 , 000g for 10min . The upper phase was gently collected ( about 600μl ) and mixed at a 1:1 ratio with 100% isopropanol and then mixed by vortexing for 2-3s . The sample was incubated for 1h at -20°C and then centrifuged ( 14 , 000rpm , 30min , at 4°C ) to collect the RNA pellet . The solution was removed without disturbing the pellet , followed by two consecutive wash rounds using 750μl 70% ethanol . The pellets were air dried for 5min and then dissolved in nuclease free H20 and incubated for 5min at 50°C . All RNA samples were treated with TURBO deoxyribonuclease ( DNase ) ( Life technologies , AM2238 ) . RNA-seq , term-seq and 5’-sequencing libraries were prepared and sequenced as previously described [33 , 48] . Sequencing was performed using the Illumina NextSeq 500 and the data was deposited in the European Nucleotide Database ( ENA ) under accession no . PRJEB21982 ( S1 Table ) . Sequencing reads generated for E . coli , S . typhimurium and E . aerogenes were mapped to the CP009273 , NC_016810 . 1 and NC_015663 . 1 RefSeq genomes , respectively , using NovoAlign ( Novocraft ) V3 . 02 . 02 with default parameters , discarding reads that were non-uniquely mapped as previously described [33] . For 5′-end sequencing , the RNA was divided into a tobacco acid pyrophosphatase ( TAP ) –treated and untreated ( noTAP ) reactions to enable primary transcript detection and then sequenced using 5’-seq [33 , 49] . TSSs were mapped as was recently described [19] . E . coli or E . aerogenes overnight cultures were diluted 1:100 into 25ml fresh LB media and incubated at 37°C until reaching an optical density ( OD600 ) of 0 . 5 . The culture was then placed in a preheated 37°C shallow water bath to preserve the experiment temperature and 125μl Rifampicin ( 100mg/ml , for a final concentration of 500μg/ml ) were immediately added to the culture to inhibit RNA synthesis . Selected time points were sampled by collecting 1 . 4ml from the culture into a pre-chilled tube containing 170ul of ice-cold stop solution ( 90% ethanol and 10% saturated phenol ) to deactivate cellular processes and RNA-decay . The sample was quickly vortexed and then placed on ice until all time points were collected . The samples were centrifuged for 5min at 4000rpm to collect cell pellets and were flash frozen . During RNA extraction , after tri-reagent mediated lysis , each sample was spiked with 5fmol of the ERCC RNA ( Ambion , 4456740 ) to allow normalization of RNA abundance between samples . RNA-seq libraries were prepared and sequenced as described above . Gene-expression was calculated as the median coverage per nucleotide ( reads/nt ) normalized by the number of reads that mapped to all ERCC spike-in RNA , an estimate which we found is more robust than the total number of reads or average coverage in cases of non-uniform decay patterns , where overlapping operon regions , or small sub-ORF regions can persist long after the full transcript has been eliminated . Transcript half-lives were calculated by fitting the decay time-course abundance measurements per gene with a delayed exponential-decay function as previously described [22] . The previously published half-lives for the 779 transcripts described in S1 Fig were taken from the “Fig 4 Source data 1” in [22] , which provides the estimated Decay rate ( λ ) per gene . To extract the half-life we calculated t1/2 = ln ( 2 ) / λ . Operon gene annotations were extracted from EcoCyc [18] ( S2 Table ) . To identify and analyze gene-pairs found within the same transcriptional unit , consecutive gene-pairs were only considered if the following criteria were met: i ) the intergenic region interspacing the genes was shorter than 200nt . ii ) The downstream gene was not associated with an independent TSS under the growth conditions of this study [19] . iii ) The upstream gene was not associated with a term-seq site displaying an intrinsic terminator signature ( i . e . , hairpin followed by a uridine stretch ) . iv ) No substantial rho-dependent termination measured in E . coli treated with the Rho inhibitor Bicyclomycin [20] . v ) The stable and unstable genes were covered by median greater than or equal to 10 and 1 reads/nt , respectively . vi ) The expression difference between the genes was not greater than 10-fold , as we find such high values were sometimes indicative of incorrect operon annotation or highly active secondary promoters . vii ) The decay rate of both genes was measured in at least one biological replicate of the experiment . Gene-pairs in which one gene was at least 2-fold more stable and 2-fold more abundant than its consecutive neighbor gene were classified as putatively decay-regulated ( S3 Table ) . We manually accepted 4 gene-pairs displaying borderline , yet consistent signal as well as 7 differentially expressed gene-pairs in which the decay rate was not measured in our experiment usually due to lack of expression in the conditions tested , but for which differential decay could clearly be identified in a recently published dataset [22] ( S3 Table ) . Term-seq libraries were prepared as previously described [33] and sequenced using a paired-end sequencing approach [48] ( S1 Table ) . The number of 3’-end reads per genomic position was calculated and for each 3′ site the average library insert length was calculated using the paired-end read mapping positions . Sites were then associated with their respective genes , requiring that the average insert length would overlap the gene coding region [48] . The position supported by the highest number of reads associated with a stabilized gene was selected as the representative 3’-terminus of the stabilized RNA in steady-state ( S4 and S6 Tables ) . In a few cases , a different , slightly less covered position was selected instead if it provided a substantially better fit for the decay pattern observed . The sequence upstream of each selected site was extracted from the genome and folded using RNAfold [35] to evaluate the predicted structure and its estimated stability ( kcal/mol ) ( S4 and S6 Tables ) . The S . typhimurium and E . aerogenes protein-coding sequences were retrieved from NCBI and blasted against the E . coli protein database , with E-value set to 10−5 . Gene and operon orthologues were classified as the Best Bi-directional Hits ( BBHs ) . Gene-pairs were compared if they occurred consecutively as in E . coli and were substantially expressed , as described above ( S4 and S5 Tables ) . Genes containing a protective structure embedded at least 50 bases into the coding region were selected ( n = 23 ) and their orthologues from up to 21 different bacterial strains belonging to the Enterobacteriaceae family ( including the E . aerogenes and S . typhimurium strains in this study ) were identified using blast ( S9 Table ) . Gene information for each organism was downloaded from the Integrated Microbial Genomes ( IMG ) database [50] . Gene orthologues were discarded if the gene sizes differed by more than 50 bases . Genes were then aligned using Clustalw2 [51] and the conservation at each position in the alignment was defined as the maximal base frequency detected at that position . Conservation was calculated for each of the codon positions independently ( as shown in S5 Fig ) . The average conservation at positions overlapping protective structures was calculated using a window size of 50 bases . In cases where the stabilizing structure occurred at the end of the gene ( as in Fig 3 ) the window was defined as 70 bases to accommodate the longer structures found in these genes . The conservation in control gene regions that do not contain a known protective structure was calculated using a sliding window approach ( using the same window size and sliding each window by 5 codons at a time over the entire protein coding region ) . WT E . coli and temperature-sensitive RNase E mutants were generously provided by the McDowall lab [37] . Triplicates of each strain were grown in the permissive temperature of 30°C overnight in LB and were diluted the next day 1:100 in fresh media . The cultures were then grown in 30°C until reaching an OD of 0 . 5 and were then incubated at 44°C for 10min to briefly deactivate RNase E in the mutant but not the WT . The cells were collected and the RNA was extracted as described above . The mono-phosphorylated 5’ends were sequenced using 5’-end sequencing ( noTAP only , as described above ) . Putative RNase E cleavage positions , tightly matching the decay pattern observed with RNA-seq , were considered if they covered by a sum of at least 20 reads across all replicates and showed >2 . 5-fold average enrichment in the WT compared with the mutant samples . In cases where multiple potential cleavage sites were available , the most highly covered site was selected . The sequence downstream of the RNase E cleavage site was analyzed and folded using RNAfold [35] ( S7 Table ) . Normalized ribosome densities ( ribosome counts per mRNA ) were retrieved from a recently published dataset [3] . The ribosome density ratio between consecutive operon gene members was calculated for all analyzed gene-pairs ( S3 Table ) in which translation efficiency could be calculated for both genes in the dataset [3] . The differential translation distributions of uniformly ( n = 533 ) and differentially decaying ( n = 39 ) gene-pairs were then compared using a two-sided Wilcoxon rank-sum test . Estimated decay rate values were retrieved from a recently published dataset , in which E . coli bacteria were pre-exposed to 1mg/ml kasugamycin for 15min before being subjected to a transcriptome-wide rifampicin-based RNA decay assay [22] . The decay ratio for the control and kasugamycin-treated samples was calculated in all cases where such values were available for both genes in at least one replicate . In cases where data was available for the two published replicates , the average decay ratio was used instead .
|
Bacteria utilize operonic transcription to synchronize the expression of multiple consecutive genes . However , this strategy lacks the ability to fine-tune the expression of specific operon members , which is often biologically important . In this report , we integrate multiple transcriptome-wide RNA-sequencing methods to show that bacteria commonly employ differential mRNA decay rates for genes residing within the same operon , generating differential transcript abundances for equally transcribed operon members , at steady state . By comparing the transcriptomes of different bacteria , we show that differential decay not only regulates the expression levels of hundreds of genes but also often evolutionarily conserved , providing support for its biological importance . By mapping the RNA termini positions at steady-state , we show that stabilized operon segments are protected from different RNases through a combination of protective RNA structures , which surprisingly , are often encoded within protein-coding regions and are evolutionarily conserved . In addition , we provide evidence that differential ribosome densities over the regulated operons guide the initial events in the differential decay mechanism . Our results highlight differential mRNA decay as a major shaping force of bacterial transcriptomes and gene regulatory programs .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"sequencing",
"techniques",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"nucleases",
"enzymes",
"pathogens",
"messenger",
"rna",
"dna-binding",
"proteins",
"operons",
"enzymology",
"microbiology",
"bacterial",
"diseases",
"enterobacteriaceae",
"genome",
"analysis",
"dna",
"molecular",
"biology",
"techniques",
"rna",
"sequencing",
"cellular",
"structures",
"and",
"organelles",
"bacteria",
"bacterial",
"pathogens",
"salmonella",
"typhimurium",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"rna",
"structure",
"genomics",
"proteins",
"medical",
"microbiology",
"ribonucleases",
"microbial",
"pathogens",
"molecular",
"biology",
"salmonella",
"ribosomes",
"biochemistry",
"rna",
"hydrolases",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"transcriptome",
"analysis",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"organisms",
"macromolecular",
"structure",
"analysis"
] |
2018
|
Extensive reshaping of bacterial operons by programmed mRNA decay
|
E-values have been the dominant statistic for protein sequence analysis for the past two decades: from identifying statistically significant local sequence alignments to evaluating matches to hidden Markov models describing protein domain families . Here we formally show that for “stratified” multiple hypothesis testing problems—that is , those in which statistical tests can be partitioned naturally—controlling the local False Discovery Rate ( lFDR ) per stratum , or partition , yields the most predictions across the data at any given threshold on the FDR or E-value over all strata combined . For the important problem of protein domain prediction , a key step in characterizing protein structure , function and evolution , we show that stratifying statistical tests by domain family yields excellent results . We develop the first FDR-estimating algorithms for domain prediction , and evaluate how well thresholds based on q-values , E-values and lFDRs perform in domain prediction using five complementary approaches for estimating empirical FDRs in this context . We show that stratified q-value thresholds substantially outperform E-values . Contradicting our theoretical results , q-values also outperform lFDRs; however , our tests reveal a small but coherent subset of domain families , biased towards models for specific repetitive patterns , for which weaknesses in random sequence models yield notably inaccurate statistical significance measures . Usage of lFDR thresholds outperform q-values for the remaining families , which have as-expected noise , suggesting that further improvements in domain predictions can be achieved with improved modeling of random sequences . Overall , our theoretical and empirical findings suggest that the use of stratified q-values and lFDRs could result in improvements in a host of structured multiple hypothesis testing problems arising in bioinformatics , including genome-wide association studies , orthology prediction , and motif scanning .
The evaluation of statistical significance is crucial in genome-wide studies , such as detecting differentially-expressed genes in microarray or proteomic studies , performing genome-wide association studies , and uncovering homologous sequences . Different biological applications have settled for different statistics to set thresholds on . In biological sequence analysis , accurate statistics for pairwise alignments and their use in database search [1–3] were introduced with the use of random sequence models and E-values two decades ago [4 , 5] . Sequence similarity searches have evolved further , from the pairwise comparison tools of FASTA [3] and BLAST [5] , to sequence-profile [6–8] and profile-profile [9–12] comparisons . While different approaches to detect sequence similarity have relied on a variety of statistics , including bit scores [13 , 14] and Z-scores [3] , most modern approaches are based on E-values . Detecting sequence similarity in order to uncover homologous relationships between proteins remains the single most powerful tool for function prediction . Many modern sequence similarity approaches are based on identifying domains , which are fundamental units of protein structure , function , and evolution . Homologous domains are grouped into “families” that may be associated with specific functions and structures , and these domain families organize protein space . Domain families are typically modeled with profile hidden Markov models ( HMMs ) [13 , 15] . There are many domain HMM databases , each providing a different focus and organization of domain space , including Pfam [14] , Superfamily [16] , and Smart [17] . Although HMM-based software , such as the state-of-the-art HMMER program [18] , has features that make it superior to its predecessors , accurate significance measures arose only recently [19] . At its core , domain prediction is a multiple hypothesis testing problem , where tens of thousands of homology models ( one for each domain ) are scored against tens of millions of sequences . Each comparison yields a score s and a p-value , defined as the probability of obtaining a score equal to or larger than s if the null hypothesis holds . While a small p-value threshold ( for example , 0 . 05 or smaller ) is acceptable to declare a single test significant , this is inappropriate for a large number of tests . Instead , thresholds for domain prediction are typically based on the E-value . The E-value can be computed from a p-value thresholds as E = pN , where N is the number of tests , and yields the expected number of false positives at this p-value threshold . E-value thresholds make sense for a single database search , especially if few positives are expected . However , E-values are less meaningful when millions of positives are obtained , and a relatively larger number of false positives might be tolerated . Moreover , in multiple database query problems , such as BLAST-based orthology prediction [20] or genome-wide domain prediction [21] , E-values are usually not valid because many searches are performed without the additional multiple hypothesis correction required . Control of the False Discovery Rate ( FDR ) is an alternative and appealing approach for multiple hypothesis testing [22] . The FDR is loosely defined as the proportion of all significant tests that are expected to be false , and can be estimated as the E-value divided by the number of predictions made . The FDR does not increase with the database size N the way the E-value does; thus , predictions do not usually lose significance with the FDR as the database grows . The FDR also does not require additional correction in the case of multiple database queries . The FDR is controlled from p-values using the Benjamini-Hochberg procedure [22] . The q-value statistic is the FDR-analog of the p-value , and it provides conservative and powerful FDR control [23] . The q-value of a statistic t is the minimum FDR incurred by declaring t significant [23] . Thus , q-values vary monotonically with p-values , and they are easily estimated from p-values [23] . While E-values control the number of false positives , q-values control their proportion . The local FDR ( lFDR ) measures the proportion of false positives in the infinitesimal vicinity of the threshold , and hence it is a “local” version of the FDR [24]; it is also equivalent to the Bayesian posterior probability that a prediction is false [24] . However , q-value estimates are much more robust than lFDR estimates , since the former are based on empirical cumulative densities , which converge uniformly to the true cumulative densities [25 , 26] . On the other hand , lFDR estimates are local fits to the density , so they are comparably more susceptible to noise , especially on the most significant tail of the distribution . The FDR [22] , q-value [23] , and lFDR [24] have all been successfully used in many areas of bioinformatics , including gene expression microarray analysis [24 , 27 , 28] , genome-wide association studies ( GWAS ) [27 , 29] , and proteomics analysis [30–34] . Here we introduce the first FDR- and lFDR-estimating algorithms for domain prediction . An essential feature of our approach is that statistical tests are stratified by domain family , rather than pooled . We prove that stratified problems are optimally tackled using the lFDR . For domain prediction , we evaluate how well thresholds based on stratified lFDRs and q-values perform using five independent approaches for estimating empirical FDRs . Through extensive benchmarking using the Pfam database and HMMER , we find that using stratified q-values increases domain predictions by 6 . 7% compared to the Standard Pfam thresholds on UniRef50 [35] . In contrast to theory , we also find that q-values outperform lFDRs . Further , while the empirical FDRs for most domain families agree with our q-value thresholds , some families tend to have larger FDRs; the standard null model appears to be inappropriate for them and yields inaccurate p-values . Specifically , families with larger-than-expected empirical FDRs are enriched for those containing repetitive patterns , such as coiled-coils , transmembrane domains , and other low-complexity regions . When only families with as-expected FDRs are considered , the use of q-values increases domain predictions by 8 . 8% compared to the Standard Pfam , and lFDRs further outperform q-values , suggesting that further performance improvements are possible if the statistical modeling of repetitive families is improved . Stratified FDR analyses have been previously explored [36–39] , and have been successfully applied to GWAS in particular [29 , 40 , 41] . Thus , the same solution we introduce for domain recognition applies to a wide variety of problems in which statistical tests can be analyzed separately , including GWAS ( stratifying by candidate or genic regions ) , orthology prediction ( stratifying by each ortholog database search ) , motif scanning ( stratifying by each motif search across a genome ) , multi-microarray analysis ( stratifying by each microarray ) , and other multi-dataset analyses . Overall , we expect the use of stratified q-values and lFDRs to yield improvements in many applications in bioinformatics and beyond .
We briefly review the relevant FDR definitions; for a comprehensive overview , see [42] . Given a p-value threshold t , let V be the number of false positive predictions , and R be the total number of significant tests . Assuming independent p-values drawn from a two-component distribution of null and alternative hypotheses ( Fig 1 ) , V and R have expected values of E[V ( t ) ]=tπ0N , E[R ( t ) ]=F ( t ) N , where π0 is the proportion of tests which are truly null , N is the total number of tests , and F ( t ) is the cumulative density of p-values [23 , 24 , 43] . Note that E[V ( t ) ] gives the E-value . There are two closely-related versions of the FDR used in our work: the positive FDR ( pFDR ) and marginal FDR ( mFDR ) [42 , 43] , defined as pFDR=E[VR|R>0] , mFDR=E[V]E[R] . The advantages of the pFDR compared to the original FDR definition of Benjamini and Hochberg [22] are discussed in [43] . If p-values are drawn independently from the two-component distribution of Fig 1 , the pFDR and mFDR were proven to be equivalent to the following posterior probability [43]: pFDR ( t ) =mFDR ( t ) =Pr ( H=0|p≤t ) =tπ0F ( t ) , where H = 0 denotes that the null hypothesis holds . This quantity is sometimes called the “Bayesian FDR” [24] . The pFDR and mFDR are also asymptotically equal under certain forms of “weak dependence , ” as defined in [44] . Our domain prediction problem has large sample sizes and weak dependence: our dataset contains millions of protein sequences and thousands of HMMs , and null p-values are only dependent for very similar sequences and similar HMMs . Dependent tests represent a very small subset of all hypotheses tested , even on each stratum ( for any one HMM ) . For this reason , we use FDR to refer loosely to all these FDR definitions . The local FDR ( lFDR ) is the Bayesian posterior error probability defined as [24] lFDR ( t ) =Pr ( H=0|p=t ) =π0f ( t ) , where f ( t ) = F' ( t ) is the p-value density at t . Thus , while the pFDR is a ratio of areas , the lFDR is a ratio of densities ( Fig 1 ) [45] . The q-value of a statistic t is the minimum pFDR incurred by declaring t significant [23] . Estimated q-values are efficiently constructed from p-values , and conservatively estimate the pFDR [23] . Specifically , q-value and lFDR estimation are based on the above formulas , where π0 , F ( t ) and f ( t ) are replaced by estimates . See the Supp . Methods in S1 Text for the algorithms for estimating q-values and lFDRs . Here we prove that the lFDR gives optimal thresholds for stratified problems . For domain prediction , each domain family defines a stratum . We wish to find p-value thresholds ti per stratum i that maximize the number of predictions across strata while constraining the maximum FDR of the strata combined . Optimality of the lFDR here is consistent with the related Bayesian classification problem , where posterior error probabilities are also optimal [43] . Let the FDR model quantities Ni , π0 , i , Fi ( ti ) and fi ( ti ) be given per stratum i . We desire to maximize the expected number of predictions across strata ∑iFi ( ti ) Ni , while constraining the “combined” FDR , which we define as the sum of expected false positives across strata divided by the total number of expected predictions , to a maximum value of Q , or ∑itiπ0 , iNi∑iFi ( ti ) Ni≤Q . This problem is solved using the Lagrangian multiplier function Λ , with the constraint set to strict equality , in a formulation that avoids quotients: Λ=∑iFi ( ti ) Ni+λ ( ∑itiπ0 , iNi−Q∑iFi ( ti ) Ni ) =∑iFi ( ti ) Ni ( 1−λQ ) +λtiπ0 , iNi . Taking the partial derivative of Λ with respect to tj , we obtain a necessary condition for optimality , ∂Λ∂tj=fj ( tj ) Nj ( 1−λQ ) +λπ0 , jNj=0⇔Q−1λ=π0 , jfj ( tj ) =lFDRj ( tj ) , which shows that the lFDR of each stratum must be equal , since the last equation has the same value for every j . Optimality of the lFDR also holds when constraining the combined E-value instead of the combined FDR ( Supp . Methods in S1 Text ) . Each of the 12 , 273 Pfam domain families was used to scan for domains in each of 3 . 8 million proteins of UniRef50 ( Supp . Methods in S1 Text ) , resulting in a total of 47 billion tests . Domain predictions are stratified by family ( HMM ) , and each stratum contains p-values from which we estimate q-values and lFDRs . We note that standard q-value and lFDR implementations fail for domain data for two reasons . First , modern HMM software only reports the smallest p-values due to heuristic filters [19] . Second , homologous families ( grouped into “superfamilies” [16] or “clans” [14] ) produce frequent overlaps that are resolved by removal of all but the most significant match , and thus there are fewer predictions than an independent family analysis would predict , which leads to underestimated FDRs . To address these issues , we remove overlapping domains ( keeping those with the smallest p-values ) , and then estimate q-values and lFDRs with methods adapted for censored p-values ( Methods ) . For comparison , we also use E-value thresholds and the “Standard Pfam” curated bitscore thresholds ( also called “Gathering” or “GA” [14] ) . Note that a stratified E-values approach ( separating families ) is no different from a combined E-value approach in that the ranking of predictions is preserved , since the number of proteins , or tests , is the same per stratum; the stratified E-value threshold equals the combined E-value threshold divided by the number of strata . Similarly , a combined q-value or lFDR approach ( obtained by combining the p-values of all strata ) also preserves the E-value rankings . We estimate the true FDR via “empirical” FDR tests , to compare all methods on an equal footing , but also to test the accuracy of q-value estimates . We created or adapted five tests , each of which labels domain predictions as either true or false positives ( TP , FP ) using different statistical and biological criteria . The proportion of predictions labeled FP estimates the FDR . For simplicity , only two tests are described here in detail and are featured in the main figures . First , the ClanOv ( “Clan Overlap” ) test is based on the expectation that overlapping domain predictions should be evolutionarily related [46] . Pfam annotates related families via clans . In this test , domain predictions are ranked by p-value , highest ranking domains are considered as TPs , domains that overlap a higher-ranking domain of the same clan are removed ( since they would not be counted as separate predictions ) , and domains that overlap a higher-ranking domain of a different clan are considered FPs ( Methods , Fig 2A ) . All FPs in this test would not be predicted by our method when overlaps are removed; nevertheless , this method estimates well the amount of noise . Second , the ContextC ( “Context Coherence” ) test is based on whether domain pairs predicted within a sequence have been observed together before [47] . For each sequence , domain predictions are ranked by p-value , and the highest ranking domain is always a TP . Subsequently , a domain is a TP if its family has previously been observed with the family of at least one higher-ranking domain , and otherwise it is a FP ( Methods , Fig 2B ) . The principles behind the other three tests are described here briefly: OrthoC ( “Ortholog Set Coherence” ) is based on the expectation that orthologous proteins contain similar domains [48] , RevSeq ( “Reverse Sequence” ) estimates noise based on domains predicted on reversed amino acid sequences [49] , and MarkovR ( “Markov Random” ) estimates noise based on domains predicted on random sequences generated from a second-order Markov model ( Supp . Methods and Fig A in S1 Text ) . Methods are compared at the same empirical FDR based on the number of domain predictions ( Fig 3 and Fig B in S1 Text ) , unique families per protein ( Fig C in S1 Text ) , amino acids covered ( Fig D in S1 Text ) , and proteins with predictions ( Fig E in S1 Text ) , as well as their total “GO information content” scores ( derived from the Gene Ontology [50] and MultiPfam2GO [51]; Supp . Methods and Fig F in S1 Text ) . Stratified q-value thresholds outperform E-values in all tests ( Fig 3 , Fig B in S1 Text ) . While stratified lFDR thresholds are superior to E-values in all tests , they are unexpectedly outperformed by q-values on most tests . We hypothesize that lFDR estimates are less robust than q-values due to errors in p-values; these errors most likely arise because of weaknesses in the standard null model . The Standard Pfam is not evaluated using ClanOv and ContextC ( Fig 3 ) ; these tests are based on the Pfam clans and observed domain pairs , so the Standard Pfam has zero empirical FDRs in both . However , q-values outperform the Standard Pfam in two of the three fair tests ( OrthoC , MarkovR ) and perform similarly in RevSeq ( Fig B in S1 Text ) . The same trends hold if the combined empirical E-value is controlled ( Supp . Methods and Fig G in S1 Text ) . We measure improvements not only of domain counts , which may be inflated for families with many small repeating units , but also of unique family counts . We also measure the information content based on the GO terms associated with domain predictions [51] ( Supp . Methods in S1 Text ) . To have amounts of noise comparable to Pfam , we calculate p- and q-value equivalents to the Standard Pam thresholds for each family ( Supp . Results in S1 Text ) . The medians of these distributions give thresholds of q ≤ 4e-4 , and for E-values , p ≤ 1 . 3e-8 ( Supp . Results in S1 Text ) . Q-values improve all metrics consistently relative to the Standard Pfam ( between 4–7% , Fig 4 ) . E-values predict 2% fewer domains than the Standard Pfam , but slightly outperform Pfam in the other metrics ( Fig 4 ) . We also evaluated dPUC , a prediction method based on domain context [48 , 52] . dPUC also improves upon the Standard Pfam in all cases ( Fig 4 ) . dPUC increases domains more than q-values , but their unique family count and amino acid coverage are comparable , and q-values best dPUC for protein counts and GO information content . This is because dPUC predicts more repeat domains ( of the same family ) and tends to restrict new predictions to proteins that already had Standard Pfam predictions . In contrast , q-values increase domains at the same rate as they increase protein coverage , which increases information the most . Thus , while stratified q-values predict fewer domains than dPUC , those domains tend to be more informative than the dPUC predictions at comparable FDRs . We find large disagreements between q-values and our empirical FDRs tests ( except for MarkovR; Fig 5 , Fig H in S1 Text ) . Interestingly , the disagreement is proportionally larger for smaller FDRs , and shrinks as the FDR grows ( Fig 5 ) . We hypothesize that a few families are too noisy at stringent thresholds , and this subset becomes proportionally smaller as all families are allowed greater noise . To test this , we compute empirical FDRs separately per family at a threshold of q ≤ 1e-2 ( Methods ) . This threshold gives a greater FDR than the Standard Pfam ( Supp . Results in S1 Text ) , which is desirable here as many families have few predictions at more stringent thresholds . Since large deviations between the empirical FDRs and q-values may arise due to low sampling , significance is assessed by modeling this random sampling ( Methods ) . We find that most families ( 92–99% , Table A in S1 Text ) have FDRs close to the q-value threshold or have statistically insignificant differences ( blue and black data in Fig 6 , Fig I in S1 Text ) . Four tests ( ClanOv , ContextC , OrthoC , and RevSeq ) detect many families with significantly larger FDRs than expected ( 3–8% , Table A in S1 Text ) . These families are significantly enriched for those containing coiled-coils , transmembrane domains , and low-complexity regions ( Fig 7; Methods ) . There are fewer families with significantly smaller FDRs than expected ( 0–2% , Table A in S1 Text ) , and they do not appear to share common patterns . Only the MarkovR test conforms to expectation , with no families having significantly larger FDRs than expected and 0 . 1% of families having significantly smaller FDRs than expected . We use the four tests ( excluding MarkovR ) to assign families into mutually-exclusive classes by majority rule . The “increased-noise” families have significantly large positive deviations ( see Methods; red in Fig 6 ) in at least three tests . The “decreased-noise” families have significantly large negative deviations ( green in Fig 6 ) in at least three tests . Lastly , the families with “as-expected-noise” have small deviations ( blue and some black in Fig 6 ) in at least three tests . There are 327 increased-noise families ( 2 . 7% of Pfam , S1 File ) , one decreased-noise family ( HemolysinCabind ) , and 4433 as-expected-noise families ( 36% , S2 File ) . There are 7512 unclassified families in Pfam ( 61% ) . Using these classes , we find that the Standard Pfam has more stringent thresholds ( in terms of q-values ) for increased-noise as compared to as-expected-noise families , but many increased-noise family thresholds remain too permissive ( Supp . Results and Fig J in S1 Text ) . Empirical FDRs agree more with q-values in as-expected-noise families than in all families combined , although some disagreement remains ( Fig K in S1 Text ) . In these families , lFDRs outperform q-values ( Fig 8 ) , as we expect from our theoretical results when the underlying p-values are correct . Compared to the Standard Pfam , domain counts at q ≤ 4e-4 increase from 6 . 7% in all families to 8 . 8% in as-expected noise families ( similar increases are observed on all metrics; Fig L in S1 Text ) , and lFDRs further improve upon q-values . Thus , lFDRs may become more useful should p-values for all families improve in the future . The previous methods describe a single “domain” threshold set via the stratified q-value or lFDR analysis . However , HMMER provides additional information in the form of “sequence” p-values , which score the presence of domain families combining the evidence of repeating domains . Only 2 . 3% families have different sequence and domain Standard Pfam thresholds [14] . Here we define “two-tier” thresholds using the FDR . In the first tier , we compute q-values from the sequence p-values and set the threshold Qseq . In the second tier , we compute q-values on the domain p-values , only for the domains in sequences that satisfied the sequence threshold , and set the threshold Qdom|seq ( corresponding to a FDR conditional on the first threshold ) . The final FDR is approximately Qseq+Qdom|seq if both thresholds are small and under an independence assumption ( Supp . Methods in S1 Text ) . For simplicity , we only evaluate the case where Qseq = Qdom|seq . Tiered q-values predict many more domains , at any fixed empirical FDR , than domain q-values and domain lFDRs , our previous two best statistics , consistently and by very large margins ( Fig B in S1 Text ) . Tiered q-values also outperform other methods in predicting new families per sequence ( Fig C in S1 Text ) ; the entire signal of these families comes from combining repeating units , none of which is significant by itself . There is also a large increase in amino acid coverage ( Fig D in S1 Text ) , and a smaller increase in protein coverage ( Fig E in S1 Text ) and GO information content ( Fig F in S1 Text ) . Tiered q-values also compare favorably to dPUC [48] , matching the superior domain improvements of dPUC , and outperforming dPUC in all other metrics ( Fig M in S1 Text ) . Thus , tiered q-values retain the strengths of domain q-values while powerfully leveraging the limited context information of repeating domains present in sequence q-values . However , the estimated FDRs of tiered q-values are less accurate than for domain q-values ( Fig H in S1 Text ) , and remain less accurate in as-expected-noise families ( Fig K in S1 Text ) . For this reason , tiered stratified q-values are experimental: although they are more powerful than domain-only q-values , they do not , as described , control the FDR as well .
In multiple hypothesis testing , the FDR and lFDR are straightforward approaches for controlling the proportion of false positives and the posterior error probability , respectively . The q-value is a statistic for controlling the FDR that is less biased and more flexible than previous FDR procedures such as the one from Benjamini and Hochberg [22] . Benchmarks based on empirical FDRs have been a part of recent works studying protein and DNA homology [47 , 48 , 52 , 53]; however , those approaches have used expensive simulations rather than estimating FDRs directly from p-values ( or E-values ) , as q-values do very efficiently . Our work is , to the best of our knowledge , the first attempt at applying q-values and lFDRs to domain identification , thus advancing the statistics of this field . Our theoretical work revealed that the lFDR , which is the Bayesian posterior probability that a prediction is false , is the optimal quantity to control in stratified problems . Stratified lFDR control has previously been found to optimize stratified thresholds in the related problem of minimizing the combined false non-discovery rate while controlling the combined FDR [37] . The lFDR also arises naturally in Bayesian classification problems [43] . Stratified lFDR thresholds ensure the least confident predictions of each stratum have the same posterior error probability . However , we found that estimated q-values are more robust than our lFDR estimates for domain predictions , where the underlying p-value estimates are imperfect [45] ( Fig 3 ) . We extended the domain stratified q-value approach into what we call tiered stratified q-values , by setting q-value thresholds on both the sequence and domain statistics reported by HMMER . While accurate FDR estimation of this procedure remains a challenge , tiered q-values successfully leverage the additional signal of repeating domains to increase predictions ( Fig M in S1 Text ) . There are other successful approaches , such as dPUC [48] and CODD [52] , that use the broader concept of domain context ( or co-occurrence ) to improve domain predictions . Remarkably , tiered q-values perform as well or better than as dPUC under all metrics ( Fig M in S1 Text ) , even though tiered q-values only utilize the context signal of repeating domains , while dPUC additionally considers context between families [48] . In the future , tiered q-values could be combined with dPUC to yield further improvements in domain prediction . We introduced a suite of empirical FDR tests to evaluate domain predictions . Altogether , these tests are powerful means for evaluating the correctness of predictions ( “Evaluation of empirical FDR tests” Supp . Results in S1 Text ) . Four of our tests consistently revealed flaws in the estimates of statistical significance for some families . We found a strong enrichment among noisy families for coiled coils , transmembrane domains , and other low-complexity regions . These problematic domain categories have been noted elsewhere [46 , 54 , 55] , and ad hoc solutions have been proposed [54 , 56] . However , none of these solutions are implemented by standard software such as BLAST and HMMER [56] . In our view , obtaining correct statistics for these repetitive families should be the top priority of the field of sequence homology . Nevertheless , most families in Pfam appear to have correct statistics , and the advantage of using q-values and lFDRs is clear . In the future , the standard sequence similarity software packages should be able to report these stratified statistics natively rather than as a post-processing step as is done here . Domain prediction is one case where stratified FDR and lFDR control are desirable , since domain families occur with vastly different frequencies and are thus associated with differing amounts of true signal . However , the same holds for other applications , such as BLAST-based orthology prediction [20] , since some ortholog groups are orders of magnitude larger than others . FDR and lFDR control may also improve iterative profile database searches , such as PSI-BLAST [6] , as well as numerous other sequence analysis problems . The basis of our work is a general theorem applicable to naturally stratified statistical tests . Whether the combined FDR or E-value is constrained , equal stratified lFDR thresholds are required to maximize predictions . Besides limits on sample size , the strata may be arbitrary , so our result can be broadly applied to multiple hypothesis testing problems . In motif scanning , for example in silico transcription factor ( TF ) binding site identification , the position weight matrix of each TF may yield a p-value per match [57] , and the number of binding sites per TF may vary by orders of magnitude across different TFs . Here , we recommend computing lFDRs stratified by TF , and setting equal lFDR thresholds across TFs . For protein domains , one could further stratify p-values using taxonomy , since domain family abundances vary greatly across the kingdoms of life ( archaea , bacteria , eukarya , and viruses ) [58 , 59] . In sum , we have demonstrated the practical utility of our theoretical contributions to domain prediction , which are likely to influence many applications in bioinformatics and beyond .
A p-value distribution is required to estimate q-values and lFDRs . HMMER reports two kinds of p-values . The “sequence” p-value combines every domain of the same family on a protein sequence , while the “domain” p-value is limited to each domain instance . The sequence p-value thus reports whether the protein sequence as a whole contains similarity to the HMM , whereas the domain p-value scores individual domain units within the sequence . We obtained domain predictions with p-values on UniRef50 [35] and OrthoMCL5 [20] proteins using hmmsearch from HMMER 3 . 0 and HMMs from Pfam 25 with these parameters: the heuristic filters “--F1 1e-1--F2 1e-1--F3 1e-2” allow sequence predictions with “stage 1/2/3” p-value thresholds of 0 . 1 , 0 . 1 , and 0 . 01 , respectively . Moreover , we obtain p-values using “-Z 1--domZ 1” . Lastly , we remove domains with p>0 . 01 by adding “-E 1e-2--domE 1e-2” . For each domain family HMM , we use its HMMER p-values over a protein database to estimate q-values and lFDRs . We use standard methods [27 , 45] adapted for censored tests since HMMER3 only reports the most significant p-values while standard methods require all p-values . Notably , HMMER3 does not provide complete p-values even if filters are removed [60] , and only small p-values are accurate [19] , so the full set of p-values is not useful . Moreover , the filters are desirable to reduce HMMER3's runtime . The Supp . Methods ( S1 Text ) reviews these standard methods for estimating q-values and lFDRs , and details our adaptations for domains . Briefly , we remove overlaps between domain predictions ranking by p-value , before computing q-values and lFDRs; otherwise , the amount of true positive may be overestimated because overlapping domains will be counted double , a common case within Pfam clans . Secondly , the standard approaches require all p-values solely to estimate π0 , here roughly the proportion of proteins that do not contain a domain family . We set π0 = 1 , which gives slightly more conservative q-values and lFDRs than otherwise . Our software for computing stratified q-values , lFDR estimates and tiered q-values from HMMER3 is DomStratStats 1 . 03 , available at https://github . com/alexviiia/DomStratStats . We compare new and standard domain prediction approaches over a range of relevant empirical FDRs . We vary thresholds based on stratified q-values and lFDRs , and compare their performances to thresholds varied by E-values and extensions of the Standard Pfam . Stratified domain E-values are computed from the HMMER p-values by multiplying them by the number of proteins in UniRef50 , as hmmsearch would compute them . The “Standard Pfam” has two expert-curated thresholds per family , for domain and sequence bitscores respectively ( Pfam calls them “gathering” thresholds ) [14] . For all methods , domain overlaps are removed ranking by p-value . Overlaps between families in the “nesting” list are not removed ( Supp . Methods in S1 Text ) . All methods use a permissive overlap definition [61] ( Supp . Methods in S1 Text ) , except for the Standard Pfam ( there overlaps of even one amino acid are removed [14] ) . The Standard Pfam thresholds are mapped to p-values , q-values , and lFDRs , and the medians of these distributions are used in comparisons ( Supp . Results and Fig N in S1 Text ) . We introduce a suite of tests that measure empirical FDRs using biologically-motivated definitions of TPs and FPs . The “standard” biological sequence null model , which most software from BLAST to HMMER use , consists of random sequences generated assuming independent and identically distributed amino acids . Domains predicted on these random sequences produce a distribution of random bit scores from which p-values are computed . The five empirical tests we use instead label every prediction as either a TP or a FP , and these labels are used to compute empirical FDRs and E-values ( number of type I errors , or FPs ) . Each test makes different assumptions , and together they provide independent and complementary evaluations . We describe our two primary tests in detail next; for the other three , see Supp . Methods ( S1 Text ) . Given domain predictions labeled as either TPs or FPs as above , we compute empirical FDRs at two levels . Briefly , the “method-level” FDR evaluates an entire scoring method ( q-values , E-values , etc . ) combining all domain families , whereas the “family-level” FDR evaluates the accuracy of q-values separately per family . These quantities are consistent estimators of the corresponding true pFDRs under weak dependence [44] . At a threshold t , let TPij ( t ) and FPij ( t ) be the observed number of true positives and false positives , respectively , for domain family j in protein sequence i . DPUC improves domain prediction by taking into account the “context , ” or presence of other domain predictions [48] . A newer version of dPUC now works with HMMER3 , among other improvements that will be described elsewhere . Context family pair counts were derived from Pfam 25 on UniProt proteins . The “candidate domain p-value threshold” of dPUC is a tunable parameter , which when set to p ≤ 1e-4 gives comparable empirical FDRs to q ≤ 4e-4 on the MarkovR and OrthoC tests . dPUC is not evaluated in ContextC because both are based on domain context ( dPUC would have a zero empirical FDR ) , nor in ClanOv because dPUC requires overlap removal while ClanOv requires observing overlaps to compute its FDR . DPUC 2 . 0 is available at http://compbio . cs . princeton . edu/dpuc . PairCoil2 [62] , TMHMM [63] , and SEG [64] , were run on UniRef50 using standard parameters to predict coiled coils , transmembrane domains , and low-complexity regions , respectively . Each Pfam family observed at least 4 times in UniRef50 was associated with a category if more than half of its domains overlapped the category's predictions . For families with multiple categories , only the one with the greatest amino acid overlap was kept . Unassigned families were categorized as “other” .
|
Despite decades of research , it remains a challenge to distinguish homologous relationships between proteins from sequence similarities arising due to chance alone . This is an increasingly important problem as sequence database sizes continue to grow , and even today many computational analyses require that the statistics of billions of sequence comparisons be assessed automatically . Here we explore statistical significance evaluation on data that is stratified—that is , naturally partitioned into subsets that may differ in their amount of signal—and find a theoretically optimal criterion for automatically setting thresholds of significance for each stratum . For the task of domain prediction , an important component of efforts to annotate protein sequences and identify remote sequence homologs , we empirically show that our stratified analysis of statistical significance greatly improves upon a combined analysis . Further , we identify weaknesses in the prevailing random sequence model for assessing statistical significance for a small subset of domain families with repetitive sequence patterns and known biological , structural , and evolutionary properties . Our theoretical findings in statistics are relevant not only for identifying protein domains , but for arbitrary stratified problems in genomics and beyond .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
Beyond the E-Value: Stratified Statistics for Protein Domain Prediction
|
We used the RIP1-Tag2 ( RT2 ) mouse model of islet cell carcinogenesis to profile the transcriptome of pancreatic neuroendocrine tumors ( PNET ) that were either non-invasive or highly invasive , seeking to identify pro- and anti-invasive molecules . Expression of multiple components of desmosomes , structures that help maintain cellular adhesion , was significantly reduced in invasive carcinomas . Genetic deletion of one of these desmosomal components , desmoplakin , resulted in increased local tumor invasion without affecting tumor growth parameters in RT2 PNETs . Expression of cadherin 1 , a component of the adherens junction adhesion complex , was maintained in these tumors despite the genetic deletion of desmoplakin . Our results demonstrate that loss of desmoplakin expression and resultant disruption of desmosomal adhesion can promote increased local tumor invasion independent of adherens junction status .
The ability of a tumor to invade into the surrounding normal tissue marks a critical step in the transition from benign to malignant tumor growth . The acquisition of this hallmark of cancer is associated with poor prognosis for many human cancers and is often considered a precursor to the development of metastases [1] . As such , considerable effort has been directed towards identifying invasion promoting and suppressing molecules and the mechanisms by which they modulate a tumor's invasive phenotype [2] . Amongst the discernible barriers to the acquisition of an invasive growth phenotype is cell-cell adhesion , and cellular alterations that result in disrupted , reduced , or otherwise functionally altered cellular adhesion are strongly associated with the progression to a malignant tumor phenotype [3]–[5] . The importance of sustaining cellular adhesion for homeostasis , particularly in epithelial tissues , is evident in the number of distinct structures whose primary function is to maintain cell-cell interconnections , which include the adherens junctions ( AJs ) , desmosomes , and tight junctions [6] , [7] . These complexes share many structural similarities , including the presence of transmembrane proteins – typified by the cadherins – that mediate adhesive connections with neighboring cells as well as intracellular molecules – exemplified by the catenin and the plakin families – that connect these transmembrane components to the cytoskeleton [6] , [7] . In particular , changes in the expression and/or function of AJ components have been associated with malignant cancers , and numerous studies have focused on the role of AJs in restricting invasive growth [3] , [8] , [9] . In this study , we utilized the RIP1-Tag2 ( RT2 ) mouse model of cancer to identify the mechanisms by which tumors acquire invasive growth capabilities . RT2 mice develop multiple pancreatic neuroendocrine tumors ( PNET ) by 12–14 weeks of age due to the expression of the SV40 T antigen oncoprotein ( Tag ) in the pancreatic β cells [10] . This model has proven useful in characterizing many aspects of tumorigenesis due to its relatively synchronous and predictable progression through distinctive lesional stages that culminate in invasive carcinomas [11]–[13] . We used this model to identify pro- and anti-invasive molecules in an unbiased fashion by comparing the non-invasive islet tumors to highly invasive carcinomas using microarray profiling of the mRNA transcriptome . We identified several components of desmosomes whose expression was significantly decreased in invasive tumors , implicating attenuation of desmosomal function in malignant progression . To assess this hypothesis , we engineered into the oncogene-expressing cancer cells in RT2 mice a genetic deletion of desmoplakin ( Dsp; MGI: 109611 ) , an intracellular protein critical for desmosomal stability [14] . Loss of Dsp led to an increased incidence of invasive carcinomas providing strong evidence that desmosomal adhesion acts as a distinct barrier to invasive tumor growth .
We chose to use the RT2 mouse model of cancer to characterize mechanisms governing the switch from benign to invasive tumor growth since a broad spectrum of invasive tumor lesions develop in end-stage RT2 animals . These include the non-invasive islet tumor ( IT ) , the focally invasive carcinoma type-1 ( IC1 ) , and the broadly invasive carcinoma type-2 ( IC2 ) [15] . To evaluate potential mechanisms regulating invasive tumor growth in this model , we isolated tissue from IT and IC2 lesions in end-stage RT2 animals by laser capture microdissection and then profiled the mRNA transcriptome . The IC2 class showed widespread transcriptional changes as compared to the IT class ( Dataset S1 ) . We chose to focus our attention on differentially expressed genes encoding components of two cell-cell adhesion structures , namely adherens junctions and desmosomes ( Table 1 ) , since elements of each were prominently downregulated . The expression of cadherin 1 ( Cdh1 , also known as E-cadherin; MGI: 88354 ) , a molecule previously demonstrated to restrict invasive growth in this and other models [8] , [16] , was decreased in IC2 lesions as expected . Interestingly , Cdh1 was the only member of AJs that was significantly altered in IC2 lesions ( Table 1 ) . In contrast , multiple genes encoding components of desmosomes were significantly reduced in IC2 lesions ( Table 1 ) . Moreover , the expression of several desmosomal genes in addition to Cdh1 was progressively reduced in the distinctive stages of PNET tumorigenesis in RT2 mice as well as in human PNETs as compared to normal human pancreatic islets , when total lesional stages , in particular ungraded tumors , were analyzed ( Figure S1 ) [13] . Although the expression of these genes was reduced in ungraded whole tumors in comparison to normal islets , their levels were further reduced in the microdissected invasive IC2 lesions ( Table 1 ) . Based on these results , we sought to determine what role desmosomal adhesion might play in regulating invasive tumor growth in this mouse model of cancer . To confirm the microarray results , we performed immunohistochemistry for multiple desmosomal components . Staining for Dsp and for one of the desmosomal cadherins , desmoglein 2 ( Dsg2; MGI: 1196466 ) , as well as for Cdh1 demonstrated that these molecules are expressed in the pancreatic islets as well as in the pancreatic ducts and the exocrine pancreas of wild-type animals ( Figure 1 and Figure S2 ) . In tumors of end-stage RT2 animals , the expression of all three molecules was maintained in IT lesions and was largely extinguished in IC2 lesions ( Figure 1 and Figure S2 ) . In contrast to Cdh1 , expression of catenin beta 1 ( Ctnnb1; MGI: 88276 ) , another component of AJs , was maintained in both IT and IC2 lesions , comparable to wild-type islets ( Figure S3 ) . This result is consistent both with the microarray result demonstrating that Cdh1 was the only AJ component to show any change in expression and with a previous study suggesting that Ctnnb1 does not contribute to RT2 tumorigenesis [17] . Collectively , these data confirm the microarray results and suggest the hypothesis that loss of desmosomal adhesion might contribute to the development of an invasive phenotype . To address the hypothesis raised by the microarray and immunohistochemistry results , we asked whether functionally disrupting desmosomal activity in vivo would promote invasive tumor growth in RT2 mice . To accomplish this , we chose to genetically delete Dsp since there is a single Dsp gene as compared to other components of desmosomes for which there are multiple non-allelic genes [6] . Furthermore , ablation of Dsp in vivo has previously been shown to impair desmosome function [14] . Since the Dsp whole body knockout is embryonic lethal [14] , we employed the Cre/loxP system to ablate the Dsp gene specifically in the pancreatic β cells , the same cells that express the Tag oncogene in RT2 mice . In combination with a DspFlox allele [18] , we used a mouse line in which a tamoxifen-regulatable Cre recombinase is controlled by the pancreatic duodenal homeobox gene 1 promoter ( Pdx1-CreER ) [19] . Pdx1 is expressed in all pancreatic lineages during development and is variably expressed in the adult pancreas , in particular being widely expressed in β cells [20] , [21] . We intercrossed RT2+; DspFlox/WT with Pdx1-CreER+; DspFlox/WT mice to generate the appropriate genotypes , and all expected genotypes and genders were observed in approximate Mendellian ratios ( Table S1 and Table S2 ) . To induce Cre activity , all Pdx1-CreER positive mice were given tamoxifen for five consecutive days beginning at 10 weeks of age when incipient tumors are first observed in RT2 mice [22] . In the absence of the RT2 transgene , genetic ablation of Dsp resulted in uniform loss of Dsp expression in the pancreatic islets , as determined by immunohistochemistry ( Figure S4 ) . Deletion of Dsp did not cause any change in Cdh1 expression or in the gross morphological appearance of the non-oncogene-expressing islets ( Figure S4 ) . Loss of Dsp was accompanied by significantly reduced Dsg2 expression in the pancreatic islets whereas the expression of insulin ( Ins ) , the hormone produced by β cells , did not appear to be affected ( Figure S5 ) . These results are consistent with compromised desmosomal adhesion , although we cannot strictly rule out the possibility that some residual desmosomal function persists in the absence of Dsp . Ablation of Dsp in normal pancreatic islets did not affect multiple physiological parameters , such as body mass and fasting glucose levels , and its expression in this tissue compartment is apparently dispensable in adult mice ( Figure S6 ) , setting the stage to assess the impact of its loss on PNETs arising from such islets . Lastly , the tamoxifen induction regimen by itself had no obvious effect on any aspect of RT2 tumorigenesis examined , including tumor invasion , when tamoxifen was applied to RT2 mice that lacked the Pdx1-CreER and DspFlox alleles ( Figure S7 ) . Induced loss of Dsp at 10 weeks of age did not affect any of the tumor growth parameters in RT2 mice that were sacrificed 4 weeks later . No significant changes were observed in the number of tumors that developed nor in the collective tumor burden when comparing RT2+; Pdx1-CreER+; DspFlox/Flox mice and littermate controls ( Figure 2A–B ) . Furthermore , the rates of tumor proliferation and tumor apoptosis , as judged by the levels of the proliferation marker Ki67 and the TUNEL assay respectively , were indistinguishable between groups ( Figure 2C–J ) . Thus , we conclude that the loss of Dsp does not affect tumor growth in this model . While conditional genetic ablation of Dsp in the angiogenic islet dysplasias and incipient solid tumors of RT2 mice had no discernible effects on tumor formation and subsequent tumor growth parameters , it did lead to an increase in tumor invasion . RT2 mice develop a spectrum of tumor lesions , including non-invasive ( IT ) , focally invasive ( IC1 ) , and broadly invasive ( IC2 ) lesions ( Figure 3A–F ) [15] . Loss of Dsp resulted in a greater frequency of invasive tumors and a concomitant reduction in the percentage of non-invasive IT tumors in mice analyzed four weeks after genetic ablation of Dsp in incipient solid tumors ( Figure 3G–H ) . Whereas ∼40% of total tumors could be classified as invasive carcinomas in control mice , greater than 60% of all tumors fell into this category in RT2+; Pdx1-CreER+; DspFlox/Flox mice ( Figure 3G ) . Interestingly , this shift appears to result from selective progression to the focally invasive IC1 but not to the widely invasive IC2 tumors . Indeed , while there is no significant change in the development of IC2 lesions ( approximately 10% of all tumors fall into this class regardless of Dsp status ) , more than 50% of tumors can be classified as IC1 lesions in RT2+; Pdx1-CreER+; DspFlox/Flox mice versus ∼30% in control mice ( Figure 3H ) . We confirmed that Dsp was in fact lost in these tumors by examining the recombination status of the Dsp allele by PCR . Tumors that were genotypically DspFlox/Flox showed near universal recombination of the Dsp allele , confirming that Dsp was lost in these tumors ( Figure 3I ) . Tumors isolated from control DspWT/WT or DspFlox/WT mice showed no recombination or were heterozygous for the recombined and wild-type Dsp alleles respectively . Thus , we conclude that the conditional genetic ablation of Dsp in incipient tumors of RT2 mice leads to increased local tumor invasion . We were intrigued that loss of Dsp led to an increase in the IC1 class but not in the IC2 class of invasive tumors . Since Cdh1 also acts as a dominant invasion suppressor in this model , we examined its status in the tumors from RT2+; Pdx1-CreER+; DspFlox/Flox mice and littermate controls by immunohistochemistry . We found that Cdh1 expression was maintained in the IT and IC1 tumors that developed regardless of Dsp status ( Figure 4I–L ) . Tumor margins and regions of invasion were identified by staining for the Tag oncoprotein ( Figure 4E–H ) . Indeed , Cdh1 appeared to be expressed at comparable levels in IT and IC1 tumor lesions regardless of Dsp status ( Figure 4M–T ) . Expression in IT and IC1 lesions of a second component of AJs , junction plakoglobin ( Jup , also known as gamma catenin; MGI: 96650 ) , was also unaffected by Dsp status ( Figure S8 ) , consistent with AJ function being maintained in these lesions despite the absence of Dsp and impaired/ablated desmosomal function . Lastly , cadherin 2 ( Cdh2 , also known as N-cadherin; MGI: 88355 ) , a marker of epithelial-mesenchymal transition ( EMT ) , was expressed at readily detectable and comparable levels in IT and IC1 tumors regardless of Dsp status , as well as in the IC2 tumors that did not express Cdh1 ( Figure S9 ) , consistent with the results of a previous study investigating determinants of progression to invasive carcinoma [8]; notably , there is no indication that activation of the invasive growth capability in this pathway involves an EMT , as reflected in differential expression of Cdh2 or other markers of EMT . Given that the expression of both Dsp and Cdh1 was lost in IC2 lesions , the most invasive class of RT2 tumors , both in unmodified RT2 mice and in tamoxifen-treated RT2+; Pdx1-CreER+; DspFlox/Flox mice ( Figure 1 and data not shown ) , we infer that loss of Dsp by itself is sufficient to promote the development of focally invasive tumors while the additional loss of Cdh1 is required to develop a more aggressive invasive tumor phenotype .
To date , much of the work on desmosomes in human disease has focused on their role in maintaining heart and skin integrity , where desmosomal defects are associated with cardiomyopathy and skin blistering conditions respectively [23] . More recently , a potential role for desmosomes in cancer progression has been suggested based on a variety of experimental clues [24] . For example , in vitro cell culture assays demonstrated that inhibiting desmosomal adhesion via blocking peptides caused morphological disorganization [25] while introduction of desmosomal components into a nonadhesive cell line resulted in increased cell aggregation and reduced cellular invasion in vitro [26] . These studies suggested that loss of desmosomal function might contribute to tumor invasion and malignancy , consistent with their role in maintaining cellular adhesion . ( Our attempts to perform similar in vitro experiments using cell lines derived from RT2 tumors [βTCs] were hindered by the fact that βTC cell lines express desmosomal components at low levels , presumably due to adaptations to culture , and generally perform poorly in migration/invasion assays – data not shown ) . In further support of the proposed role of desmosomes as a barrier to malignant progression , several pathology studies characterizing human cancers have shown that decreased or altered expression of desmosomal components , including Dsp , correlates with increased tumor invasion , advanced tumor grade , and poor patient prognosis , particularly in oral cancers where expression of desmosomal components are highly expressed in the normal oral mucosa [4] , [5] , [27] . Additionally , our bioinformatic analysis of human cancer databases confirmed that the expression of desmosomal genes is often decreased in a variety of human epithelial cancers as compared to normal tissues and is occasionally further decreased in more advanced grades of tumors ( Table S3 ) . The present study substantively extends this current state of knowledge by demonstrating that desmosomal adhesion can indeed act as a distinct barrier to the development of an invasive tumor phenotype in the in vivo setting of a genetically engineered mouse model of cancer . We identified several components of desmosomes – Dsp , Dsg2 , desmocollin 2 ( Dsc2; MGI: 103221 ) , and plakophilin 2 ( Pkp2; MGI: 1914701 ) – whose expression was significantly downregulated in the highly invasive tumor lesions that develop in the RT2 mouse model of PNET . These changes were reflected at the protein level as determined by immunostaining of non-invasive IT lesions and broadly invasive IC2 lesions . The simultaneous decrease in expression for multiple desmosomal genes suggests that there may be coordinated transcriptional regulation of desmosomal components . Prime candidates for such regulation include the transcription factors that regulate EMT , such as the Snail and Twist families of transcription factors [28] . Notably , however , we did not detect significant differential expression of such transcription factors in our microarray analysis comparing non-invasive IT and highly invasive IC2 PNETs ( Dataset S1 ) , and the expression of one prominent marker of EMT , Cdh2 , was not obviously different between IT and IC2 lesions , consistent with the results of a previous study investigating determinants of the invasive phenotype using this same model of PNET [8] . Thus , the current evidence suggests that the acquisition of an invasive phenotype in this tumor type does not involve a classical EMT . Our results clearly demonstrate that the conditional genetic deletion of a single core desmosomal component , Dsp , promotes increased local tumor invasion in RT2 mice , producing a phenocopy of such inferred transcriptional regulation in the normal circumstances of tumor progression . While desmosomes play an integral role in maintaining epithelial integrity , they are by no means the only structure involved in cellular adhesion . In addition to desmosomes , several related structures , including AJs , contribute to maintaining cell-cell adhesion [7] . However , while desmosomes and AJs play related biological roles in terms of maintaining cellular adhesion and have similar structural compositions , it is worth noting that there are clear differences in the consequences of impaired desmosome adhesion versus impaired AJ adhesion on tumor phenotypes . An elegant functional genetic study demonstrated that Cdh1 , a core member of AJs , acts as an invasion suppressor in vivo; targeting a transgene encoding a dominant-negative Cdh1 molecule to the oncogene-expressing pancreatic β cells markedly accelerated tumor progression and led to significantly increased frequencies of invasive carcinomas and to the development of lymph node metastasis in this same mouse model of PNET [8] . In comparison , deletion of Dsp led to an increase in the frequency of the focally invasive IC1 grade of islet carcinomas but not the more widely aggressive IC2 carcinomas , and distant metastases were not observed ( data not shown ) . One possible explanation for the differences in these phenotypic outcomes is the different roles that Dsp and Cdh1 play within their respective adhesion complex . While Cdh1 is a transmembrane protein that directly links cells together by forming homotypic interactions with other Cdh1 molecules on neighboring cells [29] , Dsp is an intracellular molecule that contributes to the overall stability of the desmosomal plaque and links this structure to the intermediate filaments [14] . Therefore , deletion of Dsp may attenuate but not totally abolish desmosomal function; if so , then the specific deletion of one of the desmosomal cadherins , Dsc2 or Dsg2 , might have a more pronounced effect on invasiveness . An additional explanation for the increase in the focally invasive IC1 fraction but not the broadly invasive IC2 fraction of invasive tumors following ablation of Dsp involves the observed maintenance of Cdh1 and AJs . Expression of Cdh1 as well as a second component of AJs , Jup , was retained in both the non-invasive IT tumors and in the now more prevalent focally invasive IC1 tumors following genetic deletion of Dsp . It would seem likely , in light of the aforementioned functional study in this same mouse model of cancer [8] , that the preservation of Cdh1 expression and of AJ function serves to maintain an additional , stronger brake on tumor invasion . Thus , while loss of Dsp and impairment of desmosomal adhesion leads to the focal invasion observed in IC1 lesions , the development of the broadly invasive phenotype found in IC2 lesions evidently requires the concomitant loss of Cdh1 . Indeed , the IC2 tumor lesions that normally develop in RT2 mice show a coordinated reduction in the expression of Cdh1 and multiple desmosomal components ( Table 1 , Figure 1 , and Figure S2 ) . The apparently independent regulation of desmosomal and AJ adhesion is notable since AJ stability has been proposed to affect desmosomal stability and vice versa in other contexts [18] , [30] , [31] , whereas Cdh1 and Jup are evidently not affected by the deletion of Dsp during PNET tumorigenesis in RT2 mice . Interestingly , the genetic deletion of Dsp had no consequential effects on the other parameters of RT2 tumorigenesis beyond invasion . Although it has been suggested that Dsp and other desmosomal components can affect cellular proliferation and apoptosis [32] , [33] , we did not observe any changes in tumor growth parameters following the genetic deletion of Dsp ( Figure 2 ) . Our results are consistent with one of the earliest studies to examine the role of Dsp in vivo , wherein a skin-specific deletion of catenin alpha 1 ( Ctnna1; MGI: 88274 ) , the AJ homologue of Dsp , led to increased skin proliferation and hyperplasia whereas ablation of Dsp did not [34] . Thus , with regards to the RT2 model of PNET and possibly other forms of cancer , it appears that desmosomes primarily serve to maintain cell-cell adhesion and hence suppress the acquisition of an invasive growth capability such that the observed downregulation of desmosomal genes results in the impairment of desmosomal function and a concomitant weakening in cellular adhesion without affecting other parameters of tumorigenesis . Finally , it is important to set these results into the broader context of knowledge about malignant progression to an invasive growth state in this stereotypical pathway of multistep tumorigenesis . While disrupted cell-cell adhesion caused by the reduced expression of Cdh1 [8] and/or desmosomal genes ( this report ) clearly promotes invasive tumor growth , other factors are involved as well . Thus for example , increased expression of the type-1 insulin-like growth factor receptor ( Igf1r; MGI: 96433 ) can drive these PNETs to acquire a highly invasive phenotype [15] . Additionally , the recruitment of immune cells to the margins of these PNETs has been shown to promote invasiveness , in part by supplying cathepsin proteases and heparanase ( Hpse; MGI: 1343124 ) [35]–[37] . As such , multiple factors can impact the progression to invasiveness by varying degrees ( Figure 5 ) , and future research may well identify additional components . Irrespective , our results demonstrate that loss of desmosomal adhesion , as exemplified by the genetic deletion of Dsp , can enable a tumor to acquire an invasive phenotype . The functional study presented herein establishes desmosomal adhesion as a distinct and ostensibly independent suppressor of invasive tumor growth . This knowledge will likely contribute to a better understanding of the mechanisms governing tumor progression to an invasive growth state and may prove useful in evaluating invasive states of human cancers .
All mice used in this study were housed and maintained in accordance with the University of California , San Francisco ( UCSF ) institutional guidelines governing the care of laboratory mice . The generation and characterization of the RIP1-Tag2 ( RT2 ) [10] , DspFlox [18] , and Pdx1-CreER [19] mouse lines have been previously reported . All mice were backcrossed a minimum of six generations into the C57Bl/6 ( B6 ) background ( Charles River , Wilmington , MA ) and then intercrossed to generate the specified genotypes . To induce CreER activity , mice were injected intraperitoneally with 100 µl of 10 mg/ml tamoxifen ( Sigma , St . Louis , MO ) suspended in peanut oil for five consecutive days beginning at 10 weeks of age . To relieve the effects of hypoglycemia induced by the insulin-secreting tumors , all RT2 mice received 50% sugar food ( Harlan Teklad , Madison , WI ) beginning at 10 weeks of age . Pancreata were isolated from 14-week-old mice and embedded in OCT ( Sakura Finetek , Torrance , CA ) on dry ice . Tumor number and tumor volume were quantified as previously described [12] . For histological analysis , frozen tissues were sectioned at 10 µm thickness , and every tenth section was stained with hematoxylin and eosin ( Surgipath Medical Industries , Richmond , IL ) using standard methods . Tumors were classified as a non-invasive islet tumor ( IT ) , a focally invasive carcinoma type-1 ( IC1 ) , or a broadly invasive carcinoma type-2 ( IC2 ) using a previously defined grading scheme [15] . Fresh-frozen pancreatic sections ( 10 µm ) from 14-week-old RT2 B6 mice were fixed in cold 70% ethanol for 16 hours prior to laser capture microdissection ( LCM ) . Sections were stained using a modified hematoxylin and eosin stain that preserves RNA integrity while allowing for the microscopic visualization of pancreatic structures [38] . LCM was performed using an Arcturus PixCell II laser capture microscope system ( Molecular Devices , Sunnyvale , CA ) . Total RNA was isolated using the Arcturus PicoPure RNA Isolation kit ( Molecular Devices , Sunnyvale , CA ) and DNase I treated ( Qiagen , Valencia , CA ) . Equal amounts of RNA ( 8 ng/lesion ) from three independent IT or IC2 tumor lesions were pooled , and then cDNA was generated , amplified , and biotinylated using the Ovation Biotin System ( NuGen , San Carlos , CA ) . Three independent pools per tumor class were generated for subsequent microarray analysis . Labeled cDNA was hybridized to Affymetrix Mouse Genome 430 2 . 0 arrays ( Affymetrix , Santa Clara , CA ) according to the manufacturer's specifications . Data were analyzed by the UCSF Helen Diller Family Comprehensive Cancer Center Biostatistics and Computational Biology Core . The data were normalized using a robust multi-chip averaging method utilizing the freely available R language . Linear models were fit for each pair of groups to be compared with log2 expression as the response and the tumor phenotype indicator as the independent variable using the limma package in Bioconductor . Moderated t-statistics were used , and p-values were adjusted by controlling the false discovery rate . A change in gene expression was identified as significant if the false discovery rate was less than 0 . 05 , meaning that fewer than 5% of false findings would be expected among the genes declared to be differentially expressed . Frozen tissues were sectioned at 10 µm thickness . For immunofluorescence staining , sections were fixed in cold acetone . For colorometric staining , sections were fixed in 10% Zn-buffered formalin ( Medical Chemical Corporation , Torrance , CA ) , subjected to antigen retrieval using the Antigen Unmasking Solution ( Vector Laboratories , Burlingame , CA ) , and blocked for endogenous peroxidase activity . Antibodies used in this study were as follows: rat anti-cadherin 1 ( Invitrogen , Carlsbad , CA ) ; mouse anti-desmoplakin I/II , mouse anti-desmoglein 1/2 ( Fitzgerald , Concord , MA ) ; mouse anti-catenin beta 1 , mouse anti-cadherin 2 , mouse anti-junction plakoglobin ( BD Biosciences , San Jose , CA ) ; guinea pig anti-insulin ( Millipore , Billerica , MA ) ; rabbit anti-T-antigen ( Hanahan laboratory preparation ) ; rabbit anti-Ki67 ( Novus Biologicals , Littleton , CO ) ; rhodamine red-X-conjugated donkey anti-mouse IgG , rhodamine red-X-conjugated donkey anti-rabbit IgG , FITC-conjugated donkey anti-rat IgG , FITC-conjugated donkey anti-guinea pig IgG , biotin-conjugated donkey anti-rabbit IgG ( Jackson ImmunoResearch Laboratories , West Grove , PA ) . For mouse antibodies , non-specific binding was blocked using the Mouse on Mouse Blocking Reagent ( Vector Laboratories , Burlingame , CA ) . Fluorescently labeled tissues were mounted with Vectashield mounting medium containing 4′ , 6-diamidino-2-phenylindole ( DAPI ) ( Vector Laboratories , Burlingame , CA ) to visualize cell nuclei . The TdT-mediated dUTP-digoxigenin nick-end labeling ( TUNEL ) assay was used to assess tumor apoptosis as previously described [15] . For colorometric staining , signal was amplified using the Vectastain Elite ABC kit ( Vector Laboratories , Burlingame , CA ) , visualized using Nova Red substrate ( Vector Laboratories , Burlingame , CA ) , and counterstained with hematoxylin . For Ki67 and TUNEL quantification , two to three random fields were obtained using a 40× objective lens from at least two tumors per mouse and at least five mice per group . The proliferation or apoptosis index was calculated as the percentage of total cells per field that were Ki67- or TUNEL-positive respectively using the MetaMorph software package ( Molecular Devices , Sunnyvale , CA ) . For all other immunohistochemical analysis , two to three tumors per mouse from a minimum of five mice per indicated group were analyzed per staining condition . All images were captured using an Axio Imager bright field microscope or an Axio Scope fluorescence microscope and the AxioVision LE software package ( Carl Zeiss , Thornwood , NY ) . Fisher's exact test and the chi-square test were used to compare tumor invasion metrics . The Mann-Whitney test was used to compare tumor burden , tumor number , tumor proliferation rates , tumor apoptosis rates , and body mass metrics . The Mann-Whitney and the Wilcoxon matched pairs test was used to compare fasting glucose metrics . For all statistical tests , a p-value of p≤0 . 05 was considered significant . All statistics were performed using the Prism software package ( GraphPad Software , La Jolla , CA ) . Animals were fasted overnight for 14–16 hours prior to the first tamoxifen injection and one week following the final tamoxifen injection . Fasting glucose levels were measured using a FreeStyle Freedom glucose meter ( Abbott Laboratories , Abbott Park , IL ) . Tumor tissue was isolated directly from OCT embedded tissues , and genomic DNA was isolated using the QIAmp DNA Micro kit ( Qiagen , Valencia , CA ) . PCR was performed using standard methods . Primers used were as follows: Cre ( forward: 5′-CATGTTCAGGGATCGCCAGG-3′ and reverse: 5′-TGCGGTGCTAACCAGCGTTTT-3′ ) ; β2 microglobulin ( forward: 5′-CACCGGAGAATGGGAAGCCGAA-3′ and reverse: 5′-TCCACACAGATGGAGCGTCCAG-3′ ) ; Dsp-WT/Flox ( forward: 5′-GGTTGGGCCTCTCGAATCATGAGTGTCTAGCG-3′ and reverse: 5′-TGTCTGTTGCCATGTGATGCC-3′ ) ; Dsp-Recombined/Non-Recombined ( forward: 5′-ACAGGCCAGATGAGATCACC-3′ and reverse: 5′-TGTCTGTTGCCATGTGATGCC-3′ ) . Normal islets were isolated from six-week-old wild-type B6 mice , and hyperplastic islets were isolated from six-week-old RT2 B6 mice as previously described [39] . Angiogenic islets were isolated from nine-week-old RT2 B6 mice by selection based on their red , hemorrhagic appearance following collagenase digestion of pancreata [39] . Islet tumors were excised from the surrounding exocrine pancreas from 14-week-old RT2 B6 mice . Total RNA was purified using the RNeasy Mini kit ( Qiagen , Valencia , CA ) and DNase I treated ( Qiagen , Valencia , CA ) . cDNA was synthesized using the iScript cDNA Synthesis kit ( Bio-Rad Laboratories , Hercules , CA ) . Real-time quantitative PCR was performed using a 7900HT system ( Applied Biosystems , Foster City , CA ) ( see Table S4 for a complete list of primers used in this study ) according to the manufacturer's specifications .
|
The ability of a tumor to invade into the surrounding normal tissue is one hallmark of a malignant cancer . We sought to identify factors that either restrict or promote tumor invasion in a genetically engineered mouse model of pancreatic neuroendocrine cancer by characterizing the transcriptional profiles of the non-invasive and invasive pancreatic neuroendocrine tumors ( PNET ) that develop in this model . This analysis demonstrated that multiple genes encoding components of desmosomes , cellular structures dedicated to the maintenance of cell-cell adhesion , were expressed at much lower levels in invasive PNETs , suggesting that loss of desmosomal adhesion contributes to the development of an invasive phenotype in these tumors . Genetic deletion of one of these desmosomal components in PNET-bearing mice resulted in increased local tumor invasion . These results are important since the development of an invasive phenotype is associated with a poor prognosis for many human cancers and is often a precursor to the development of distant metastases . Our findings demonstrate one mechanism by which tumors can acquire such an invasive phenotype and may prove useful in evaluating the malignancy of human cancers .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"oncology",
"genetics",
"and",
"genomics/animal",
"genetics",
"genetics",
"and",
"genomics/gene",
"discovery",
"genetics",
"and",
"genomics/disease",
"models",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/gene",
"function",
"genetics",
"and",
"genomics/cancer",
"genetics"
] |
2010
|
Genetic Deletion of the Desmosomal Component Desmoplakin Promotes Tumor Microinvasion in a Mouse Model of Pancreatic Neuroendocrine Carcinogenesis
|
The centromeric regions of all Saccharomyces cerevisiae chromosomes are found in early replicating domains , a property conserved among centromeres in fungi and some higher eukaryotes . Surprisingly , little is known about the biological significance or the mechanism of early centromere replication; however , the extensive conservation suggests that it is important for chromosome maintenance . Do centromeres ensure their early replication by promoting early activation of nearby origins , or have they migrated over evolutionary time to reside in early replicating regions ? In Candida albicans , a neocentromere contains an early firing origin , supporting the first hypothesis but not addressing whether the new origin is intrinsically early firing or whether the centromere influences replication time . Because the activation time of individual origins is not an intrinsic property of S . cerevisiae origins , but is influenced by surrounding sequences , we sought to test the hypothesis that centromeres influence replication time by moving a centromere to a late replication domain . We used a modified Meselson-Stahl density transfer assay to measure the kinetics of replication for regions of chromosome XIV in which either the functional centromere or a point-mutated version had been moved near origins that reside in a late replication region . We show that a functional centromere acts in cis over a distance as great as 19 kb to advance the initiation time of origins . Our results constitute a direct link between establishment of the kinetochore and the replication initiation machinery , and suggest that the proposed higher-order structure of the pericentric chromatin influences replication initiation .
Centromere function , the ability to assemble a kinetochore and mediate chromosome segregation during meiosis and mitosis , is critical for the survival and propagation of eukaryotic organisms . Defects in centromere/kinetochore complexes lead to genome instability and susceptibility to cancer and cell death [1]–[3] . Therefore , it is important that properly functioning centromeres be established on both sister chromatids following replication of centromeric DNA and prior to the initiation of chromosome segregation . Centromeres in the budding yeast Saccharomyces cerevisiae replicate early in S-phase [4]–[8] and increasing evidence suggests that early centromere replication is conserved among fungi and is prevalent for at least a subset of centromeres in higher eukaryotes [9]–[13] . Yet surprisingly little is known about the mechanism that accomplishes this early replication . It has been hypothesized that early replication of centromere DNA provides sufficient time for the centromere-specific histone , CenH3 , to be incorporated on both sister chromatids and to ensure subsequent microtubule attachment [14]–[16] . Consistent with this idea , Feng et al . showed that a delay in centromere DNA replication in the absence of the replication checkpoint leads to increased aneuploidy [16] . It is thought that this observed increase in aneuploidy is due to the lack of properly bi-oriented sister chromatids . Currently , it is unclear if centromeres play an active role in their own early replication by influencing activation of nearby origins of replication ( origins ) . An alternative possibility is that centromeres have migrated over evolutionary time to reside in early replicating regions of the genome . Indirect evidence supporting the idea that centromeres play an active role in their own early replication comes from a study examining epigenetic inheritance of centromeres in the pathogenic yeast Candida albicans [11] . The centromeres of C . albicans are considered regional centromeres akin to those of higher eukaryotes because they are not defined by any distinct DNA sequence . Instead , regional centromeres are defined by broad stretches of repetitive DNA sequences ranging from about 3 kilobases ( kb ) in C . albicans to megabases in humans . Regional centromeres that have been analyzed for replication time have been shown to contain origins within them [9] , [10] , [17] . Koren et al . showed that the de novo formation of an early activated origin within the neocentromeric region is responsible for the early replication of spontaneously formed neocentromeres in C . albicans [11] . The authors further showed that the origin recognition complex ( ORC ) , which is essential for origin activation , is recruited to the neocentromere . Thus , centromeres of C . albicans appear to recruit at least a subset of the required replication initiation machinery . However , it is unclear from these studies whether centromere function directly influences origin activation time or if these regions merely provide a favorable environment for ORC recruitment , with early origin activation time being determined independently of centromere function . Distinguishing between these possibilities is difficult because no distinct sequences for either centromeres or origins have been identified in C . albicans . Therefore , the function of the centromere responsible for replication initiation cannot effectively be separated from the kinetochore-binding portion of the centromere . For example , removal of centromere DNA in C . albicans results in removal of the origins contained within that sequence . Therefore , determining if centromeres regulate origin activation time requires a situation in which the activity of the origin responsible for centromere replication can be separated from the centromere . The small sizes and well-defined sequences of centromeres and origins in S . cerevisiae provide us the opportunity to answer this question . The origins of S . cerevisiae have been extensively studied and are defined by an 11 base pair ( bp ) consensus sequence that is necessary but not sufficient to initiate DNA synthesis [18] . Unlike C . albicans , the centromeres in S . cerevisiae are small , spanning only 125 bps and do not , with one possible exception ( CEN3 ) , contain potential origins [6] , [19]–[21] . However , all of the centromeres of S . cerevisiae reside in early replicating portions of the genome , suggesting that centromeres play a role in regulating origin activation time . Early replication is not restricted to portions of the genome immediately flanking centromeres . In fact , early and late replicating blocks of DNA are interspersed throughout the genome [7] , [8] . The patterns in which temporal blocks are arranged indicate that origins can be crudely grouped into four classes with respect to their replication time and chromosomal position: centromere-proximal early , noncentromere-proximal early , telomere-proximal late , and nontelomere-proximal late [6]–[8] . Telomeres delay the activation times of nearby origins while late activation of non-telomeric origins is determined by an unknown mechanism involving unspecified DNA sequences located up to 14 kb away from the origin [22] , [23] . Plasmids containing the minimum required sequence for origin activation tend to replicate early , suggesting that early activation is the default state for origins [24] . However , in a plasmid construct , a DNA element immediately downstream of the URA3 gene can advance origin activation time of an adjacent copy of ARS1 through unknown means [25] , [26] . In light of these data , we hypothesize that there is no common “default” activation time and that centromeres act as one of the determinants for early origin activation . By measuring replication time locally as well as genome-wide in a strain with an ectopic centromere , we show that centromeres in S . cerevisiae act in cis to promote early activation on origins positioned as far as 19 kb away . Our results suggest that centromere-dependent early origin activation has a gradient effect such that the closer the origin is to the centromere the more profound is the timing effect on that origin . In addition , we show for the first time that early activation of centromere proximal origins is dependent on centromere function , suggesting that the ability of centromeres to establish proper kinetochore-to-microtubule attachments is important for regulating origin initiation .
S . cerevisiae centromeres lie close to origins that undergo replication in early S-phase . To test the hypothesis that centromeres contribute to the early activation of these adjacent origins , we obtained a strain in which the centromere on chromosome XIV ( Figure 1A ) had been relocated to a distal location on the left chromosomal arm [27] ( Figure 1B ) . At this new location the centromere is positioned in a late replicating region near the potential origin ARS1410 . We reasoned that if centromeres can influence origin activation time , then the origins that are closest to the moved centromeres would have the greatest chance of being affected . The replication kinetics of three loci on chromosome XIV ( met2 or MET2 , ARS1410 , and ARS1426 ) in the rearranged and wild type ( WT ) strains were examined using a modified version of the Meselson-Stahl density transfer experiment [4] . Haploid cells grown in the presence of dense 13C and 15N isotopes were arrested prior to S-phase . The cells were synchronously released into medium containing isotopically light carbon and nitrogen , and samples were collected at various times during the ensuing S-phase . Newly synthesized DNA was composed of light isotopes resulting in replicated DNA being hybrid or heavy-light ( HL ) in density whereas unreplicated DNA remained heavy-heavy ( HH ) in density . DNA was extracted from each cell sample , digested with restriction enzyme EcoRI , and subjected to ultracentrifugation in cesium chloride gradients . The gradients were then drip fractionated , and the kinetics with which the EcoRI restriction fragments containing each of the three loci of interest shifted from heavy to hybrid density were compared via slot blot analysis ( Figure 2A; also see Materials and Methods ) . WT cells entered S-phase by 40 minutes after release from alpha factor arrest , and most of the cells reached 2C DNA content between 120 and 140 minutes ( Figure 2B ) . The percent replication of MET2 , ARS1410 , and ARS1426 genomic fragments was calculated for each sample and plotted with respect to time ( Figure 2C ) . The time of replication ( Trep ) for each locus was calculated as the time it reached half maximal replication ( see Materials and Methods ) . ARS306 , one of the earliest known origins , and R11 , a late replicating fragment on chromosome V , were used as timing standards for comparison . To facilitate comparison between cultures , these Trep values were converted to replication indices [22] by assigning ARS306 a replication index ( RI ) of 0 and R11 an RI of 1 . 0 . Most other genomic loci have RIs between 0 and 1 . 0 . The Trep values for MET2 , ARS1410 , and ARS1426 were then converted to RIs corresponding to the fraction of the ARS306-R11 interval elapsed when the Trep for each locus was obtained . As previously observed , MET2 replicated late in the WT strain ( RI = 0 . 87 ) , as did its nearest ARS , ARS1410 ( RI = 0 . 77 ) , while ARS1426 replicated early ( RI = 0 . 16 ) ( Figure 2C and 2D ) . Similar to WT cells , the cells with the relocated centromere entered S-phase by 40 minutes following release from alpha factor arrest and reached 2C DNA content between 120 and 140 minutes ( Figure 2B ) . In contrast to the WT strain , both met2 and ARS1410 replicated early with respective RIs of 0 . 24 and 0 . 23 ( Figure 2C and 2D ) . Consistent with ARS1410 being the origin from which met2 replicates , ARS1410 maintained a slight timing advantage over met2 . Meanwhile , ARS1426 became later replicating ( RI = 0 . 79 ) in the absence of its nearby centromere ( Figure 2C and 2D ) . Similar results were obtained using an independent segregant ( see Figure S1 ) . There are three explanations for the change in replication time of the met2 locus: ( 1 ) the centromere advances the time of activation of ARS1410; ( 2 ) the centromere increases the efficiency ( percentage of cells in which an origin is active ) of ARS1410; and/or ( 3 ) insertion of the centromere created a new origin at the site . To distinguish among these possibilities we examined the replication of ARS1410 , ARS1426 , and met2 or MET2 by two-dimensional ( 2D ) gel electrophoresis [28] . The presence of bubble-arcs indicated that ARS1410 is indeed a functional origin in both the WT and rearranged strains ( Figure 3A ) . Highly efficient origins display a more intense bubble-arc signal relative to the Y-arc [29] . Based on the similarity of bubble- to Y-arc ratios we conclude that the centromere has no obvious effect on the efficiency of ARS1410 ( Figure 3A ) . A similar result was obtained for ARS1426 ( Figure 3A ) supporting the idea that the observed timing changes are not due to changes in efficiency of existing origins . No bubble-arc was observed for the MET2 locus in the WT strain or the met2 locus in the rearranged strain ( Figure 3B ) , indicating that an origin was not created by insertion of the centromere into the MET2 locus . Together these data suggest that the presence of a nearby centromere or some feature of its pericentric DNA induces early activation of origins . DNA sequences that determine origin activation time have remained largely elusive . However , previous work has shown that sequences flanking a subset of origins on chromosome XIV delay the activation of those origins [22] . These sequences , coined “delay elements” , can reside up to 14 kb from their affected origins . Therefore , it is possible that by integrating the centromere into the MET2 locus in the rearranged strain , a delay element responsible for making ARS1410 late activating was disrupted or pushed out of its effective range , thereby causing the origin to fire early . Although this scenario would explain the change in the replication times of met2 and ARS1410 , the observation that ARS1426 became later replicating when the centromere was removed from its endogenous position argues in favor of the timing changes being a consequence of centromere proximity . Alternatively , it is conceivable that there is an uncharacterized sequence element , distinct from centromeric sequence , residing in pericentric DNA that is promoting early activation of nearby origins . The existence of such an element is formally possible as a cryptic sequence residing at the 3′ end of the URA3 gene was shown to advance the activation time of nearby origins on a plasmid and in an artificial chromosomal setting [25] , [26] . We tested these possibilities as described below . At the sequence level , S . cerevisiae centromeres are composed of three essential elements: CDEI , CDEII , and CDEIII . CDEI and CDEIII are defined by essential consensus sequences whereas CDEII is a 78–86 bp AT-rich sequence that separates CDEI and CDEIII [1] , [19] , [20] . CDEIII has been found to be the element most important for centromere function as it is the binding site for essential inner kinetochore proteins , notably members of the CBF3 complex [1] , [19] , [30] . To ask if a functional centromere is required for early activation of nearby origins , we engineered a strain to have a non-functional centromere with a mutated CDEIII motif integrated at the MET2 locus while the functional centromere remained in its endogenous position ( Figure 4A ) . This strain was also subjected to flow cytometry ( Figure 4B ) and replication timing analysis as described above ( Figure 4C ) . Unlike the dramatic replication timing change observed when we introduced a functional centromere , introducing the mutated centromere caused no replication timing change at met2 and ARS1410 ( RI of 0 . 81 and 0 . 74 , respectively; Figure 4D ) . Therefore , we conclude that centromere function is needed to effect a timing change on nearby origins . Furthermore , the late replication of this region was not due to inactivation of ARS1410 through insertion of the mutated centromere as indicated by 2D gel analysis ( Figure 4E ) . Together , these data demonstrate that functional centromeres actively advance the activation times of origins over a distance of 11 . 5 kb . Upon finding that centromeres advance the activation time of origins to a distance of at least 11 . 5 kb , we sought to determine how far the centromere's effect extends along the chromosome . Raghuraman et al . presented statistical evidence that the regions of chromosomes within 25 kb of centromeres replicated earlier than the genome average , raising the possibility that centromeres could influence origin activation time over this larger distance [6] . Interestingly , at least one early replicating origin can be found within 12 . 8 kb of every centromere [16] , suggesting that centromeres may be able to influence the activation time of origins over at least this distance . A recent study using an S-phase cyclin mutant [7] showed that early replicating domains that include a centromere can be well over 100 kb , implying that the range over which centromeres can regulate origin activation time might be quite broad . To determine the range over which a centromere can influence replication time we performed a genome wide analysis of replication in the WT and rearranged strains . Cells were grown in dense medium ( see Materials and Methods ) and timed samples were collected following release into S-phase in light medium . To obtain a genome-wide view , the HH and HL DNAs from each timed sample were labeled with different fluorophores , cohybridized to microarray slides , and replication profiles were generated ( Figure 5A , Figures S2 and S3 ) . Peak locations in the profile correspond to the locations of origins while the timed sample in which the peaks first appear gives an indication of the time at which the corresponding origins become active during S-phase [6] . Replication profiles generated for the WT strain ( Figure S2 ) were consistent with previous studies [8] . Chromosome XIV displayed a strong peak at ARS1426 at the earliest time ( 40 minutes ) in the time course and a less well-defined peak at ARS1410 ( Figure 5A ) . Conversely , replication profiles in the rearranged strain displayed a shallow and late appearing peak at ARS1426 while the peak at ARS1410 appeared strong and early ( Figure 5A and Figure S3 ) confirming the observations made by slot blot analysis of individual restriction fragments . To directly compare replication profiles from the two strains , the percent replication values from the 40 , 45 , and 65 minute samples were normalized by conversion to Z-scores ( see Materials and Methods ) and superimposed on the same axes ( Figure 5B; Figures S4 , S5 , and S6 ) . Comparison of the Z-scores on chromosome XIV ( Figure 5B ) indicates that centromeres have a drastic influence over the activation times of their closest origins , suggesting that centromeres , mechanistically , operate locally . Z-score profiles for WT and rearranged chromosomes display a prominent early appearing peak centered about 19 kb to the left of the endogenous centromere . ARS1425 is a potential origin located between ARS1424 and the endogenous centromere [31] . However , 2D gel analysis demonstrated that ARS1424 is the origin likely responsible for this peak as ARS1425 is not active in either strain ( data not shown ) . Replication timing analysis of ARS1424 by slot blot hybridization indicates that this origin is influenced by the centromere to a far lesser extent than ARS1426 ( WT RI = 0 . 11 , Rearranged RI = 0 . 27 ) . That this mild effect on ARS1424 is not reflected in the Z-score overlays is likely due to the higher resolution of slot blot analysis compared to microarray analysis . We then asked if the moved centromere influenced the activation times of origins located on other chromosomes . The replication kinetics of ARS1 , a well-characterized origin on chromosome IV , were examined by slot blot hybridization and found to be similar between the two strains ( WT RI = 0 . 68 , Rearranged RI = 0 . 66 ) indicating that the activation time of ARS1 is not influenced by centromere position on chromosome XIV . The remaining profiles were examined by Z-score comparisons looking for possible trans effects of centromere position . The profiles were strikingly similar ( Figures S4 , S5 , and S6 ) . Only one other location , corresponding to ARS1531 on chromosome XV , showed a timing difference as drastic as that seen for ARS1410 and ARS1426 ( Figure 6A; Figures S4 , S5 , S6 ) . A prominent peak of high percent replication corresponding to ARS1531 was present in the WT strain while no peak was observed in the rearranged strain . However , further examination using 2D gel electrophoresis revealed that although ARS1531 is not an active origin in the isolate of the rearranged strain used for microarray analysis it is active in the WT strain as well as in the rearranged independent segregant used in the slot blot analysis of timing ( Figure 6B; see Figure S1 for kinetic data for the independent segregant ) . These data suggest that the apparent timing change at this chromosome location is likely to be a consequence of a polymorphism affecting the ability of ARS1531 to function as an origin rather than from any long-range effect of the centromere . Sequencing of ARS1531 revealed that this origin in strain YTP16 contains a mutation in the essential ARS consensus sequence ( ACS ) [31] converting it from ATATTTATATTTAGA to ATACTTATATTTAGA ( Figure 6B ) . A change of the T to a C at this position has been shown to entirely abolish origin activity in ARS307 [32] . ARS1531 does not contain this substitution in either YTP12 or YTP15 . To confirm that this basepair change is responsible for the observed difference in origin activity , ARS1531 was cloned from YTP12 and YTP16 and tested for its ability to produce transformants in YTP12 , YTP16 , and YTP19 . The “T” version of ARS1531 resulted in the formation of many robust colonies in the three strains tested whereas the “C” version of ARS1531 did not ( data not shown ) .
In this study we investigated the long-standing question of why centromeres replicate early in S-phase . We considered two possibilities: ( 1 ) that some component required for centromere function is also involved in early origin activation , or ( 2 ) that evolution has favored the migration of centromeres to early replicating regions . It has been hypothesized that early origin initiation in S . cerevisiae is the default state for origin timing [24] , suggesting a more passive mechanism for early centromere replication . However , the observation that early centromere replication is conserved [9]–[11] in conjunction with the identification of a DNA element that is not associated with centromeres but capable of advancing origin activation time [25] , [26] indicates that establishment of early origin activation time is more complex than previously thought . Consistent with this idea , a recent study shows that the centromeres of C . albicans can alter the replication times of the loci in which they reside by allowing the formation of a de novo early firing origin [11] . Because the neocentromere in C . albicans creates a new origin , it is unclear whether C . albicans centromeres also directly influence origin activation time . These results raised the question of whether the relocation of a centromere would have a similar effect in organisms with point centromeres . In this study , we took advantage of the well-characterized centromeres and origins in S . cerevisiae to effectively separate the centromere from origin function and address these questions . We show that centromeres in S . cerevisiae are capable of advancing the replication time of genomic regions in which they reside by inducing early activation of their adjacent origins at distances of 11 . 5 and 6 . 8 kb ( compare ARS1410 and ARS1426 in WT and rearranged strains , Figure 2C and 2D ) . We also show that early activation of ARS1410 and ARS1426 depends on centromere function . We find that centromere-mediated early origin activation requires an intact CDEIII region , suggesting that early origin activation is dependent on at least some portion of the DNA-protein complex normally formed at the centromere . Thus , centromeres and at least a subset of the kinetochore proteins they assemble participate as cis-acting regulatory elements of origin firing time . Furthermore , our 2D gel results indicate that centromeres do not affect origin efficiency , suggesting that the mechanisms responsible for centromere-mediated early origin activation are distinct from those that determine efficiency . S . cerevisiae centromeres are known to reside in large early replicating portions of the genome spanning as much as 100 kb and containing multiple origins [5]–[8] . In light of the finding that centromeres regulate the activation times of their closest origins , we were interested in determining over what distance centromeres exert their effect . Our genome-wide replication timing analysis indicates that the centromere's influence on origin activation time is severely diminished at a distance of ∼19 kb ( Figure 5B and 5C ) indicating that not all early replicating origins are under the centromere's influence . This result implies that there are at least two distinct mechanisms by which origins can fire early . In contrast to what has been observed in C . albicans , we see no evidence that the centromeres of S . cerevisiae create new origins . The genome is not randomly organized within the nucleus but particular genomic regions co-localize or cluster into functional foci during processes such as DNA replication [33] , [34] . In particular , centromeres in S . cerevisiae cluster throughout the cell cycle [35] . Therefore , it is plausible that centromeres , their neighboring origins , as well as other portions of the genome that interact with them , are clustered in G1 phase when timing decisions are made . This clustering could provide a way for centromeres to mediate replication time through a trans-acting mechanism . To determine if the mechanism responsible for centromere mediated early origin activation is capable of acting in trans , we examined the genome wide replication timing data for other timing changes occurring in the genome as a result of the repositioned centromere on chromosome XIV . We looked for differences in the Z-score profiles between matched S-phase samples , demanding that they persist over the course of S-phase ( See Figures S4 , S5 , S6 ) to be considered significant . Upon inspection , only one other location centered on ARS1531 displayed a timing difference of at least the same magnitude as observed for ARS1410 or ARS1426 ( compare Figure 6A; Figures S4 , S5 , S6 ) . 2D gel analysis and sequencing indicate that this timing difference is not a result of relocating the centromere position on chromosome XIV and that it is likely due to a polymorphism affecting the ability of ARS1531 to function in this particular isolate . Other than these three locations , the replication profiles are remarkably similar , suggesting that the mechanisms by which centromeres influence origin activation time are restricted to relatively limited adjacent regions . However , it is possible that centromeres contribute to smaller timing differences observed on other chromosomes such as that observed on chromosome XV at ∼850 kb ( Figure 6A; Figures S4 , S5 , S6 ) . Because the timing change at this location was only present in two of the three Z-scored samples , we did not consider it to be a significant timing difference . Here we show direct evidence of a molecular link between the establishment of the kinetochore and replication initiation machinery . Although the mechanism of centromere-mediated early origin activation is unknown , we show that such a mechanism is dependent on at least some of the protein components associated with the kinetochore that require an intact CDEIII region . We also show that the mechanism is capable of affecting initiation time of origins that reside up to ∼19 kb away from the centromere . We propose four possible models , which are not mutually exclusive ( Figure 7 ) : ( 1 ) The nuclear environment near the microtubule organizing center ( MTOC ) is particularly enriched in replication initiation factors; ( 2 ) The tension exerted by the microtubule is translated along the nearby DNA , altering its chromatin structure , thereby influencing the accessibility of the imbedded origins to initiation factors; ( 3 ) Proteins within the kinetochore directly ( or indirectly ) interact with initiation factors , recruiting them to nearby origins; and ( 4 ) The C-loop architecture of the pericentric chromatin ( see below ) ensures the origins within the C-loop will be at the periphery of the chromatin mass and are therefore more exposed to initiation factors . Studies examining the concentration of replication initiation proteins near MTOCs have not been conducted; however , nuclear pore complexes are enriched near MTOCs in S . cerevisiae [36] , [37] , the significance of which is unknown . While it is tempting to invoke localization to the nuclear periphery in the vicinity of MTOCs as a potential link between replication timing and centromeres , there is as yet no clear causal connection between nuclear localization and replication timing . Late origins tend to dwell at the nuclear periphery while early origins tend to be internal to the nucleus [38] . However , tethering an early origin to the nuclear periphery is not sufficient to alter its replication time [39] and conversely , delocalizing telomeres from the nuclear periphery does not advance their replication time [40] . Although experiments that test for protein-protein , protein-DNA , and DNA-DNA interactions of kinetochore/centromeric components have been reported , we are not aware of any experiments that bear directly on the second model ( tension-mediated promotion of early origin activation ) . Kinetochores are multiprotein complexes composed of over 60 proteins [19] , [41] . Some of these proteins have been shown to interact both genetically and physically with proteins involved in regulating DNA replication [42]–[44] . However , as we are interested in molecular components that specify origin initiation time and these components have yet to be identified , it is not straightforward to determine which interacting protein candidates should be singled out for further study . A systematic way to parse through the various kinetochore components would be to determine if members of the inner- , mid- , or outer-kinetochore complexes are required for early origin initiation through the use of temperature sensitive alleles . Though this method poses it own set of challenges , it may prove to be fruitful in uncovering some of the mechanisms by which centromeres regulate origin activation time . In vitro studies of protein-DNA interactions have shown that the CBF3 complex ( composed of inner kinetochore DNA binding proteins ) induces a 55° bend in centromeric DNA [45] . In concordance with this finding , more recent work from the Bloom lab has shown that pericentric DNA adopts a particular tertiary structure coined the C-loop [46] . The C-loop is characterized by an intramolecular loop centered on the centromere resembling a hairpin of duplex DNA . C-loop formation is dependent on a subset of inner kinetochore proteins that physically bind centromere DNA , namely those that bind the CDEIII region [46] , [47] . Cohesin , which is enriched in pericentromeric regions [48]–[50] , is thought to stabilize the C-loop following centromere replication [46] . Additonally , this study reports that the C-loop is present in G1-phase during alpha factor arrest , suggesting that neither the presence of a sister chromatid nor pericentric cohesin is essential for the maintenance of the structure . Interestingly , the intramolecular interaction of the C-loop is reported to extend more than ∼11 . 5 kb and less than 25 kb from the centromere , a distance similar to the findings reported in this study over which centromeres can regulate origin activation time . Anderson et al . hypothesized that proteins involved in the C-loop formation provide an essential geometry required for centromere position and accessibility for microtubule binding [47] . In light of these data , it is tempting to hypothesize that this geometry provides a favorable environment for early origin initiation by increasing the accessibility of nearby origins to replication initiation machinery . Although the DNA binding components of the inner kinetochore are not well conserved , in human cells a DNA binding component of the inner kinetochore , CENP-B , has been shown to induce a ∼59° bend in its target sequence [51] , suggesting that regional centromeres also adopt a particular configuration . Therefore , this model can also explain how centromeres can vary greatly at the sequence level yet still affect timing of their associated origins .
Strain and vector genotypes can be found in Table S1 . Primer sequences can be found in Table S2 . All S . cerevisiae strains used in this study were derived from a cross of CH1870 obtained from [27] and S288C . CH1870 is a mating type alpha strain derived from the S288C background . The MET2 locus of this strain ( located on chromosome XIV ) was disrupted by an insertion of centromere VII sequence and LEU2 . The endogenous centromere on chromosome XIV of CH1870 was also replaced with a URA3 selectable marker [27] . The ura3-52 allele in YTP13 ( derived from a CH1870 to S288C cross ) was restored to URA3 by gene replacement , resulting in YTP15 ( WT strain in this study , Table S1 ) . Restoration was confirmed by Southern blot analysis . Independent segregants of the rearranged strains ( YTP12 and YTP16 ) were obtained from separate tetrads from a CH1870 to S288C cross as met− , leu+ , and ura+ progeny . Three different PCR reactions using primer pairs 71∶72 , 72∶88 , and 71∶126 were used to confirm that these strains contained the desired inserted sequence at the MET2 locus ( data not shown ) . The desired insertion was further confirmed by Southern blot analysis on genomic DNA that was digested with XbaI ( data not shown ) . YTP19 was constructed to have the same met2::CEN7 . LEU2 cassette as YTP12 and YTP16 save for a 3 bp mutation in CDEIII . The mutation of CDEIII in the centromere sequence integrated at MET2 was made non-functional through site directed mutagenesis [52] . This strain was constructed as follows: Two halves of met2::CEN7 . LEU2 were PCR amplified from YTP12 using primer pairs 71∶133 , and 72∶134 . Primer pair 71∶133 was used to amplify the 5′ end of met2 with primer 71 hybridizing upstream of the BsmI restriction site located in met2 and primer 133 hybridizing downstream of the EcoRV restriction site located in the LEU2 gene . Primer pair 72∶134 was used to amplify the 3′ end of met2 with primer 72 hybridizing downstream of met2 and primer 134 hybridizing upstream of the EcoRV restriction site located in the LEU2 gene . The PCR products were sequentially digested with either BsmI or AatII followed by an EcoRV digest . The two fragments were then inserted into a pUC18 vector that contained KanMX and ARS228 through a tri-molecular ligation reaction , creating plasmid pTP18 . The centromere on plasmid pTP18 was made non-functional , creating pTP19 , through site directed mutagenesis using primers 147 and 148 [52] . These primers were used to mutate the Centromere DNA Element III ( CDEIII ) from TCCGAA to TCTAGA , introducing an XbaI site into the sequence . The most important bases for centromere function are the middle CG in TCCGAA [1] , [45] . The mutated centromere sequence on pTP19 was Sanger sequenced to ensure that only CGA had been changed . The lack of centromere activity in pTP19 was confirmed by a plasmid stability assay in which serially diluted cells were spotted to selective medium over 24 generations ( data not shown ) . The 4025 bp AatII/BsmI fragment of plasmid pTP19 containing met2::cen7 . LEU2 was transformed into YTP15 [53] , [54] . Transformants were selected on synthetic dropout ( SD ) medium plates lacking leucine . Cells were then replica-plated to SD medium lacking methionine . Colonies that were prototrophic for leucine and auxotrophic for methionine were further screened similar to the rearranged strains mentioned earlier using PCR primer pairs 71∶72 and 72∶88 and through Southern blot analysis on XbaI digested genomic DNA . Origin activity was analyzed by standard 2D agarose gel electrophoresis techniques performed on total genomic DNA obtained from either asynchronous or synchronous S phase cells [55]–[57] . Asynchronous samples were used for 2D gels conducted on ARS1426 , ARS1531 , and met2::CEN7 . Synchronized samples were used for 2D gels conducted on ARS1410 and MET2 . For asynchronous samples , cells were collected in early log phase . For synchronized samples , cells were arrested in G1-phase with alpha factor ( final concentration 3 µM ) and released synchronously into S-phase by the addition of pronase ( 0 . 15 mg/mL ) . Cells were then collected every 2 minutes and pooled . Genomic DNA was harvested similar to asynchronous samples . In the first dimension , DNA was separated in 0 . 4% agarose for 18–20 hours at 1 V/cm . The second dimension was run for 3–3 . 5 hours in 1 . 1% agarose containing Ethidium Bromide ( 0 . 3 µg/mL ) at ∼5–6 V/cm at 4°C . Cells were harvested by mixing with 0 . 1% sodium azide in 0 . 2 M EDTA and then fixed with 70% ethanol . Flow cytometry was performed as previously described [58] upon staining cells with Sytox Green ( Molecular Probes , Eugene , OR ) . The data were analyzed with CellQuest software ( Becton-Dickinson , Franklin Lakes , NJ ) . Density transfer experiments were performed as described in [4] with slight modifications in cell synchronization and sample preparation for microarray analysis [6] . Cells were grown overnight at 23°C in a 5 mL culture of dense medium containing 13C-glucose at 0 . 1% ( w/v ) and 15N-ammonium sulfate at 0 . 01% ( w/v ) . Cells were then diluted into a larger vessel in dense medium and allowed to reach an optical density of 0 . 16 ( ∼2×106 cells/mL ) . Cells were arrested in G1 by incubating with 3 µM alpha factor until at least 95% of cells were unbudded based on microscopic analysis . Cells were then filtered and transferred to medium containing 12C-glucose ( 2% ) , 14N-ammonium sulfate ( 0 . 5% ) , and alpha factor . Cells were then synchronously released from the G1 arrest through the addition of pronase at a concentration of 0 . 15 mg/mL . Samples were collected every 10 minutes . Cell samples were treated with a mixture of 0 . 1% odium azide and 0 . 2 M EDTA then pelleted and frozen at −20°C . Genomic DNA was extracted from pelleted cells , digested with EcoRI , and the DNA fragments were separated based on density by ultracentrifugation in cesium chloride gradients . Gradients were drip fractionated and slot blotted to nylon membrane where they were hybridized with probes of interest . Unreplicated DNA is HH in density while replicated DNA is HL in density . To construct replication kinetic curves , slot blots were hybridized to sequences of interest , and the percent replication in each timed sample was calculated as described in [4] , [6] , [8] . A detailed protocol can be found at http://fangman-brewer . genetics . washington . edu/density_transfer . html . The time at which the kinetic curves reach half maximal defines the time of replication ( Trep ) for each individual locus within the population of cycling cells . However , because not all G1 cells will enter or complete S-phase , Trep for the probed regions in this study were calculated based on the plateau of the replication kinetic curve for a genomic probe consisting of EcoRI digested total genomic DNA from a strain lacking mitochondrial DNA . To compare timing differences between strains , Trep values were converted to replication indices in the following manner . First the Trep values of two timing standards , ARS306 ( a known early replicating region ) and R11 ( a known late replicating region near ARS501 ) , were obtained and assigned the values of 0 and 1 , respectively . Next the Trep values of regions of interest were assigned a replication index corresponding to the fraction of the ARS306-R11 interval elapsed at the time at which the Trep for each locus was obtained . Replication index is calculated as ( Trep ( X ) −Trep ( ARS306 ) ) / ( Trep ( ARS306 ) −Trep ( R11 ) ) where X is the fragment of interest . For microarray analysis , slot blots were hybridized with a genomic DNA probe to indentify fractions containing either HH or HL DNA . Once identified , the HH and HL fractions were separately pooled and differentially labeled with cyanine ( Cy3 or Cy5 ) conjugated dUTP ( Perkin Elmer ) [8] . Timed samples were hybridized to high density Agilent yeast ChIP-to-chip 4×44 K arrays according to the manufacturer's recommendations . The algorithms for analyzing microarray data are described in [16] . An 18 kb window was used for overall smoothing . To facilitate direct comparison of replication profiles , percent replication values obtained from microarray analysis were converted to Z-score values . Z-score values were calculated using the following formula: Z = ( X−μ ) /σ where X = the percent replication value for a given probe , μ = the genomic average percent replication for a given sample , and σ = the standard deviation of the distribution of X . The 55 minute samples were not used because the genomic percent replication values were not well matched between the two strains . Two recent findings in yeast add interesting perspectives on the centromere effect on origin activation . First , the lack of forkhead proteins Fkh1 and Fkh2 results in delayed activation of a large number of normally early-firing origins , but centromere-proximal origins remain early-firing ( Knott et al . 2012 , Cell 148: 99–111 ) . Second , while many origins show delayed activation in meiotic S phase compared to mitotic S , centromere-proximal origins do not show such a delay ( Blitzblau et al . , PLoS Genetics , in press ) . Both findings are consistent with the idea that some aspect of centromere function directly influences the activation of centromere-proximal origins through a mechanism that is independent of other , more global controls .
|
Genome duplication requires the orderly initiation of DNA synthesis at sites called origins of replication . It has long been known that different origins become active at different times in S-phase ( the period during which cells duplicate their chromosomes ) . Although such temporal regulation of replication is broadly conserved among eukaryotes , how this regional control of replication time occurs largely remains a mystery . The early replication of baker's yeast centromeres ( genetic elements essential for proper segregation of chromosomes during cell division ) is one frequently cited example of temporal regulation , yet the biological significance of early centromere replication also remains speculative . Increasing evidence suggests that early centromere replication is a conserved feature of the DNA replication program across many species . Here , we show that centromeres in this yeast can advance the time at which origins in their genomic neighborhood initiate DNA replication . The distance over which centromeres can influence origin activation time extends up to 19 kilobases . We further show that centromere-mediated early origin activation depends on the centromere's ability to recruit at least a subset of the proteins needed for chromosome segregation . This study thus provides the first direct functional link between kinetochore establishment and the mechanisms of DNA replication initiation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"chromosome",
"structure",
"and",
"function",
"centromeres",
"microbiology",
"model",
"organisms",
"dna",
"replication",
"molecular",
"genetics",
"dna",
"chromosome",
"biology",
"biology",
"molecular",
"biology",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"genomics",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] |
2012
|
Functional Centromeres Determine the Activation Time of Pericentric Origins of DNA Replication in Saccharomyces cerevisiae
|
Precise patterning of morphogen molecules and their accurate reading out are of key importance in embryonic development . Recent experiments have visualized distributions of proteins in developing embryos and shown that the gradient of concentration of Bicoid morphogen in Drosophila embryos is established rapidly after fertilization and remains stable through syncytial mitoses . This stable Bicoid gradient is read out in a precise way to distribute Hunchback with small fluctuations in each embryo and in a reproducible way , with small embryo-to-embryo fluctuation . The mechanisms of such stable , precise , and reproducible patterning through noisy cellular processes , however , still remain mysterious . To address these issues , here we develop the one- and three-dimensional stochastic models of the early Drosophila embryo . The simulated results show that the fluctuation in expression of the hunchback gene is dominated by the random arrival of Bicoid at the hunchback enhancer . Slow diffusion of Hunchback protein , however , averages out this intense fluctuation , leading to the precise patterning of distribution of Hunchback without loss of sharpness of the boundary of its distribution . The coordinated rates of diffusion and transport of input Bicoid and output Hunchback play decisive roles in suppressing fluctuations arising from the dynamical structure change in embryos and those arising from the random diffusion of molecules , and give rise to the stable , precise , and reproducible patterning of Bicoid and Hunchback distributions .
Pattern formation of multicellular organisms requires the accurate spatial regulation of gene expression . Such regulation has been attributed to the position dependent distribution of signaling molecules called morphogens [1] , [2] . One of the best studied morphogens is Bicoid ( Bcd ) in embryos of Drosophila melanogaster , which works as a transcription factor to regulate the expression of hunchback ( hb ) and other downstream genes [3]–[5] . Bcd is synthesized from the maternal mRNA which is localized at around the anterior pole of embryo , and the concentration of Bcd exhibits a gradient with approximately exponential decay with distance from the anterior pole [6] . Although the decisive importance of the gradient in concentration of Bcd has been established [7] , [8] , the mechanism for forming the stable Bcd gradient and that for the reliable readout of the gradient remain largely unresolved . Important advances have been made by visualizing spatial patterns of distributions of the input protein Bcd and the output protein Hunchback ( Hb ) with the antibody staining technique [6] , [9] , [10] . These visualized data , however , have been taken from fixed embryos , which provided only the snapshots of developing embryos . Further insights were gained by combining the method using the enhanced green fluorescent protein ( eGFP ) with the antibody staining technique: In two papers [11] , [12] , Gregor et al . reported the data of dynamically changing Bcd patterns in living embryos by observing fluorescence of a Bcd-eGFP fusion protein . These data provoked fundamental questions as to the formation and readout of the Bcd gradient , which presents challenges in computational biology . In their first paper [11] , Gregor et al . showed that the Bcd distribution is approximately an exponential gradient in intranuclear concentration , which is established rather rapidly in less than 90 min after fertilization . As Bcd is a DNA-binding protein , Bcd is localized in nuclei . However , by observing responses of distributions of Bcd-eGFP to the photo-bleaching stimuli , it was shown that Bcd is not simply trapped in nuclei but is in dynamic equilibrium between influx and efflux with the cytoplasm and intranuclear degradation . The early embryo in the blastodermal stage undergoes a series of syncytial mitoses from nuclear cycles 10 to 14 , during which the nuclei are not separated by cell membranes . It was observed that Bcd which is stored in nuclei is released into cytoplasm as nuclei lose their envelopes during each mitosis , and the released Bcd is concentrated into nuclei again as envelopes are reformed during the interphase period . These observations raised the question on how the stable Bcd gradient is established in spite of the repeated dynamical structural change of embryo due to mitoses of nuclei . Though the effects of dynamically changing nuclei have begun to be to be investigated theoretically [13]–[15] , most of the hitherto developed theoretical models of the Bcd gradient have been based on the assumptions of diffusion and degradation of Bcd in a static continuum [16]–[20] . In the second paper [12] , the authors reported the data of fluctuations in concentration profiles of Bcd and Hb and showed that formation of those profiles is precise and reproducible: It was shown that fluctuations of profiles of Bcd and Hb in each embryo are about 10% , which are small enough to distinguish adjacent nuclei , and that fluctuations from embryo to embryo are also as small as 10% . These data imply that the system exerts precise control over concentration of Bcd , and that Hb is produced as a reliable response to the small concentration difference of Bcd . Response of the hb activity to Bcd , however , is the process to receive randomly arriving Bcd molecules at the hb enhancer on DNA . The random arrival of Bcd at the nanometer scale region of DNA should induce intense noise in the hb activity and Gregor et al . argued that this effect should be so large that we have to assume some unconfirmed mechanism working to reduce the noise of the readout process to the observed level of precision . There are further sources of intrinsic noise in the readout process arising from the bursting production of proteins and the nonadiabatic gene switching [21]–[28] , so that the observed precise readout of the Bcd gradient implies that the noise arising from those sources is also suppressed to the level of 10% fluctuations . In this paper , we construct two computational models , one-dimensional and three-dimensional models of Drosophila embryo , to discuss two major issues raised by Gregor et al . ; how the stable patterning of the Bcd gradient is realized in the dynamically changing syncytial embryo; and how the Bcd gradient is read in a precise and reproducible way in spite of intense noise induced by the random diffusion of Bcd molecules . Our one-dimensional model of embryo focuses on the latter issue on the noise induced by random diffusion , and effects of the dynamically changing embryonic structure are examined with the three-dimensional model . Importance of the noise induced by random diffusion has been stressed by Berg and Purcell in their early study of bacterial chemotaxis [29] . They have argued that certain physical limit of sensitivity should exist when the bacterial receptor receives the randomly arriving chemical signals . Bialek and colleagues [30]–[32] extended this analysis in a theoretically more systematic way and showed that the physical limit arising from random diffusion of signaling molecules is not the special feature of bacterial chemotaxis , but is common to many biological processes of receiving chemical signals . Gregor et al . [12] applied Bialek's theory to the problem of Drosophila embryo and showed that the limit of accuracy in sensing the Bcd concentration is ( 1 ) where c is the intranuclear concentration of Bcd , and δc is fluctuation in the Bcd concentration sensed by the hb enhancer . The hb enhancer is assumed to have the typical length a . D is the intranuclear diffusion constant of Bcd , and τ is the time length during which Bcd can bind to the enhancer . In Drosophila embryo , decision of whether hb is activated or not is made at the point of threshold concentration ; hb is activated in the nuclei of and is kept silent if . Since the observed values of D and are and [12] , and the size of the enhancer binding site should be as small as , Equation 1 implies at the decision point of the hb activation [12] . When we regard τ as the typical time length of one nuclear cycle , we have , which leads to . In Ref . 12 , δc was estimated from the data of fluctuation of the Hb concentration by assuming that δc is reflected to the hb gene activity and hence is reflected to the fluctuation in concentration of Hb . The value observed in this way was , which is much smaller than the expected value of . To explain the small value of , τ should be almost two hours , which is unacceptably long and near to the entire time available for development from fertilization up to cellularization . Thus , some mechanism lacking in this argument is needed to reduce the expected large noise of to the level of the observed precision . We may refer to this problem as the paradox of signal interpretation . Though Bialek and colleagues derived Equation 1 by solving the coupled equations of diffusion , degradation , and binding/unbinding processes , the same result can be obtained by the intuitive argument as illustrated in Figure 1: Since Bcd molecules diffuse over a distance during the allowed time τ , Bcd molecules interacting with the target of size a during the period τ are distributed in the volume . We call this volume “interaction volume” . The number of molecules in the interaction volume is in average . If Bcd molecules diffuse as independent particles , fluctuation in n should obey the Poisson statistics to exhibit . Then , we have , which gives the lower limit of fluctuation , and in more generic cases with other additional sources of noise , we have the criterion of Equation 1 . In the following , we explicitly use the idea of interaction volume to model the readout process in embryo . In the next section , we first discuss the one-dimensional stochastic model of embryo . To focus on the problem of the paradox of signal interpretation , we neglect the dynamical structural change of embryo and assume that nuclei stay still without exhibiting nuclear cycles . With this model we perform the stochastic simulation of reactions and diffusion of Bcd and Hb . Bcd molecules are produced at the anterior pole , diffuse , degrade , and bind/unbind to/from the hb enhancer , and Hb molecules are produced with the rate depending on the hb enhancer status , and diffuse and degrade . With this one-dimensional model , fluctuations in distributions of Bcd and Hb can be examined in a wide range of parameters and we show that the self-averaging process due to the Hb diffusion should resolve the paradox of signal interpretation . Effects of the dynamically changing embryonic structure are investigated by constructing a more realistic three-dimensional stochastic model of embryo . In the three-dimensional model , an embryo is represented by a cylinder consisting of 43 , 750 hexagonal sites . Bcd molecules are synthesized at the most front slice of this cylinder . The inside structure of this cylinder is modified during simulation by following the rules governing nuclear mitoses in the model: At the initial instance , one site at the core of the cylinder is assumed to be a nucleus . This nucleus starts to divide , and at the nuclear cycle 8–10 , the multiplicated nuclei move to the peripheral . At the nuclear cycle 14 , about 6000 sites at the surface layer of the cylinder are nuclear sites . Within the nuclear site the same reactions are assumed as in the one-dimensional model , and in other cytosolic sites degradation and diffusion of Bcd and Hb are considered . We show that patterning of the Bcd profile is stable and precise in spite of dynamic mitoses of nuclei in embryo . We further confirm that the same mechanism of self-averaging of Hb molecules as discussed with the one-dimensional model also works in the three-dimensional model to resolve the paradox of signal interpretation .
In Figure 3A calculated with the one-dimensional model with the standard parameterization is plotted as a function of position , . Here , results of 100 simulation runs are superposed and their average is compared with the observed data ( Figure 5A of Ref . 12 ) . The panel shows that adjustment of a parameter J enables the model to consistently reproduce the observed Bcd profile in a semi-quantitative way . Plotted in Figure 3B is for 100 simulation runs and their average . Plotted in Figure 3C is , which measures the relative gene activity at the ith site , and its average . and are plotted in Figure 3D and and are plotted in Figure 3E . We see that fluctuations in and are much more significant than in and as expected from the smallness of and . These large fluctuations are responsible for the large fluctuation of shown in Figure 3C . In spite of such large fluctuation in , fluctuation in seems to be moderate . Fluctuation in can be quantitatively analyzed with the input/output ( IO ) relation by regarding Bcd as input and Hb as output . We define the mean input as , where ihalf is the position satisfying and is the maximal value of in the system . The mean output is defined by . Thus defined mean IO relation is plotted in Figure 4A , which agrees well with the experimental data ( Figure 4A of Ref . 12 ) . Fluctuation in the IO relation is shown in Figure 4B , where we plot as a function of with . We find that the amplitude of fluctuation shown in Figure 4B is about 10% , which is as small as was observed in experiment ( Figure 4B in Ref . 12 ) . The fluctuation of output data shown in Figure 4B can be converted to the form of equivalent input data as in Figure 4C by using the relation , where , and . We can compare with of Equation 1 to see whether the output fluctuation is as large as that expected in Bialek's argument: As plotted in Figure 4C , the calculated fluctuation in output , , is around 10% , which is as small as was observed in experiment ( Figure 4C of Ref . 12 ) and is much smaller than the value expected in the argument of Equation 1 . Thus , the present one-dimensional model shows that either the random diffusion of Bcd and Hb or the noisy expression of hb does not bring about large fluctuations in the IO relation . It should be noted that in the one-dimensional model the IO relation is calculated by averaging quantities over the ensemble which is different from that in the experimental data: In experiment , many stripes along the AP axis of a single embryo can be used as an ensemble of data to measure the fluctuation of Bcd or Hb concentration in single embryo , and the small IO fluctuation observed in this ensemble is denoted by “preciseness” in Ref . 12 . In the present one-dimensional model , many runs of simulation are necessary to obtain the ensemble of enough size , and we here regard the smallness of fluctuations in this ensemble of many simulation runs as “preciseness” in the readout process . The same ensemble as in the experimental measurement can be taken in the three-dimensional model , and we will show later that the results of the three-dimensional model are essentially same as those of the present one-dimensional model , which should validate the use of this ensemble in the one-dimensional model to analyze precision in the readout process . Effects of the parameter variation on the precision can be examined by changing parameters one by one from values in the standard parameterization . In Figure 5 , the IO relation is analyzed by changing parameters , Vr , fb , fh , Nburst , and Dh . Parameters fb and fh define the rate of change in the hb enhancer status . Instead of using the bare values of fb and fh , we here use the normalized forms of ωb = fb/kb and ωh = fh/kh to explain the results . From computational [25] , [28] and theoretical [24] investigations , we can expect that decrease in the value of ωb or ωh leads to increase in noise in the Hb synthesis . The large Nburst can also be a source of noise in the Hb synthesis [22] . Results of Figure 5 indicate , however , that changes in these parameters do not strongly affect the IO relation unless too small ωb or ωh is used . When Vr or Dh is varied , on the other hand , the clear change of the IO relation can be found: Smaller Vr gives the larger noise in the IO relation , showing that the random diffusion of Bcd and Hb are dominant sources to limit the precision of readout as was suggested by Gregor et al . [12] . Dh = 0 corresponds to the case that the spatial heterogeneity of hb expression is directly reflected in the Hb distribution . In this case the fluctuation in the IO relation is large as was expected from the argument of Gregor et al . [12]: The random diffusion and random reception at the hb enhancer lead to the fluctuation larger than the observed data , which exceeds the required precision to distinguish neighboring nuclei . These dependences on parameters are further examined in Figure 6 . In Figure 6A , sensitive dependence of the fluctuation of Hb concentration on Vr and Dh is shown . When we assume the non-diffusive Hb with Dh = 0 , the calculated results show the small enough amplitude of fluctuation only when Vr is larger than 10−2 . With the realistic value of , small but finite Dh with Dh is necessary to explain the observed 10–20% fluctuation in the readout process . Effects of changing ωb and ωh were also examined ( Figure S1 ) : Small enough values of ωb and ωh bring about the large fluctuation , but in the parameter region of ωb>1 and ωh>1 , the fluctuation amplitude is moderate and does not much depend on ωb or ωh . In Figures 6B and 6C , dependence of the profile of Hb distribution on Vr and Dh is examined . The number of Hb molecules become very small when Vr≪10−3 and Dh>0 . 5 µm2/s , but does not sensitively depend on Vr or Dh when Vr>10−3 or Dh<0 . 5 µm2/s ( Figure 6B ) . The angle of gradient of Hb distribution , on the other hand , sensitively depends on Dh ( Figure 6C ) . From this dependence of the angle on Dh , we see that Dh should be less than 0 . 3 µm2/s to explain the observed experimental profile of Hb distribution [12] . Diffusion constant of Dh<0 . 3 µm2/s is consistent with the data that Hb does not diffuse over the long distance during a nuclear cycle [36] . Sensitive dependence of the fluctuation of Hb distribution on Vr is the result expected from the argument of Gregor et al . [12] and Bialek et al . [30]–[32] , but the suppression of fluctuation by nonzero Dh is rather unexpected . To understand the reason why the small but finite Dh suppresses the Hb fluctuation , we performed simulation by using Equation 8 , the assumption of disability of Hb to bind to DNA . With this assumption , the positive feedback loop of the hb regulation is lost and the cooperativity at binding sites of the hb enhancer is reduced . By putting hh = 0 and prohibiting the state S = 3 , the number of produced Hb decreases to about 25% of the amount calculated with the standard parameterization as was observed in the experiment of Ref . 40 . The slope angle of the Hb concentration at the threshold position becomes slightly smaller due to the decrease of cooperativity in the hb activation , which is consistent with the observation of Ref . 36 . The features of fluctuation in the IO relation , however , do not show a significant difference from the results obtained with the standard parameterization . In Figure 7 , sensitive dependence of fluctuation in the IO relation on Vr and Dh in the case of hh = 0 is shown . The fluctuation is dominated by the small Vr effect , but is substantially suppressed by the small but finite diffusion constant of Hb with 0<Dh<0 . 3 µm2/s . Thus , we conclude that not the feedback loop of the hb regulation but the self-averaging due to the diffusion of Hb is sufficient to suppress the large fluctuations in the hb expression . With Dh≈0 . 1 µm2/s , Hb molecules synthesized at around each nucleus are averaged over positions of several neighbor nuclei through diffusion . This simple mechanism is sufficient to suppress the fluctuation in the Hb distribution and hence resolves the paradox of signal interpretation . Shown in Figure 9 are snapshots of a simulation run with the three-dimensional model , in which the Bcd concentration is represented by green shaded colors . As the mitoses are repeated from nuclear cycle 10 to 14 , the number of nuclei on the surface of embryo increases and the cortical layer becomes thick . Bcd is concentrated in nuclei which can be recognized as the bright green dots and is also accumulated in the cortical layer to exhibit the green shaded region . These features are similar to the observed snapshots in distribution of Bcd-eGFP ( Figure 6A of Ref . 11 ) . Profiles of distribution of Bcd and Hb and the distribution of the hb enhancer state are examined in Figure 10 . In the left panels of Figure 10 ( Figures 10A , 10D , 10G ) , results of 10 simulated embryos are superposed , each of which is obtained by monitoring nuclear sites along a stripe of the cylinder surface at an instance in the interphase period of nuclear cycle 14 . In the middle panels of Figure 10 ( Figures 10B , 10E , 10H ) , the embryo-to-embryo fluctuations of the Bcd distribution , , the Hb distribution , , and the hb enhancer state , are plotted , where and is the ensemble average over 10 embryos . We can see from Figure 10 that the embryo-to-embryo fluctuation of Bcd distribution is as small as 10% as in the observed data ( Figure 5B of Ref . 12 ) . , , and can be transformed to the positional fluctuations as , , and , which are plotted in the right panels of Figure 10 ( Figures 10C , 10F , 10I ) . Positional fluctuation of Bcd , , is comparable with the observed data ( Figure 5C of Ref . 12 ) showing that the Bcd gradient is generated in a reproducible way with small embryo-to-embryo fluctuation at the interphase of nuclear cycle 14 . Fluctuations of the Hb distribution ( Figure 10E ) and the hb enhancer state ( Figure 10H ) are larger than fluctuation of Bcd ( Figure 10B ) ; At the threshold position of x/L∼0 . 48 , and amount to 40–50% . The positional fluctuations , and , however , are less than 5% ( Figures 10F and 10I ) because of the sharp change in and at the threshold position . Thus , Bcd , Hb , and the hb enhancer state are reproducible with the small embryo-to-embryo fluctuation . We calculated the IO relation by taking the average over many nuclei in a single embryo at the interphase of nuclear cycle 14 . Lines of the IO relation obtained from each of 10 runs of simulation are superposed in Figure 11A . These 10 lines do not deviate much from each other , which is the feature consistent with the experimental data ( Figure 4A of Ref . 12 ) . Fluctuations in the IO relation are quantified in Figures 11B and 11C , which show a good agreement with the observed data ( Figures 4B and 4C of Ref . 12 ) . Small values of each IO fluctuation shown by each line of Figures 11B and 11C imply that the Bcd gradient is read in a precise way in each embryo , and the small deviation of multiple IO lines from each other implies that the Bcd gradient is read out in a reproducible way with the small embryo-to-embryo fluctuation . In this way , results of the three-dimensional model show that both the generation of the Bcd gradient and the readout of the Bcd gradient are precise and reproducible at the nuclear cycle 14 in spite of the large dynamical structure change of embryo through nuclear cycle 1–14 . Dependence of the IO relation on parameters has also similar features to those discussed with the one-dimensional model ( Figure S2 ) . The fluctuation of the IO relation , σh at nuclear cycle 14 , does not sensitively depend on ωb , ωh or Nburst as far as ωb or ωh is not extremely small . The IO fluctuation is sensitive , on the other hand , to Vr and Dh . Thus , the qualitatively same results as the one-dimensional model were obtained with the three-dimensional model , which supports the view that the fluctuation in the IO relation is dominated by the small value of Vr , and the fluctuation is masked by the small but finite Dh . When the positive feedback of Hb is turned off by posing hh = 0 , the amplitude of Nh decreases as in the one-dimensional model , but features of preciseness and reproducibility are not altered much ( Figure S3 ) , which implies that the positive feedback mechanism is not necessary to suppress the fluctuations . This mechanism to resolve the paradox of signal interpretation can be visually confirmed by taking snapshots of the side-view of the model embryo ( Figure S4 ) . The hb enhancer state , γ , shows large spatial heterogeneity at the threshold region with about half of nuclei being randomly activated to produce Hb molecules for both two cases of Dh = 0 . 1 µm2/s ( Figure S4A ) and Dh = 0 ( Figure S4B ) . When Dh = 0 . 1 µm2 s−1 , thus produced Hb diffuses over several nuclei , which smoothes out this heterogeneity to produce a boundary of the Hb distribution ( Figure S4C ) , which is in sharp contrary to the case of Dh = 0 for which the Hb distribution directly reflects the heterogeneity of the gene activity ( Figure S4D ) . Though hb mRNA is not explicitly taken into account in the present model , we may be able to expect that the noise level of the hb mRNA distribution is in between the noise level of distribution of the hb gene activity and that of the Hb distribution since mRNA first localizes in nuclei in which the hb gene is active but gradually diffuses out of nuclei to represent both characteristics of the localized hb gene activity and the more delocalized diffusive hb products . This expectation is consistent with the observed larger fluctuation of the hb mRNA distribution than the Hb distribution ( Figure S8 of Ref . 36 ) . With the three-dimensional model , the simulated temporal change of the Bcd distribution during nuclear cycle 10–14 can be compared with the experimental data . In Figure 12A , the simulated Bcd concentration at the anterior part is compared with the experimental data [11] . Plotted are numbers of Bcd molecules in nuclear and cortical sites at the outmost layer of the cylinder in the slice i = 10 ( x/L = 0 . 14 ) . The Bcd concentration in nuclei is large during the interphase and sharply drops to the smaller values during mitoses . During the mitotic period , Bcd is released to the cortical layer and the Bcd concentration in cortical sites increases . In the observed data , the Bcd concentration in nuclei sharply rises in the beginning of the interphase period and then gradually decreases during the interphase period . Gregor et al . showed that this decrease in the Bcd concentration should be attributed to increase in size of each nucleus during the interphase period: With increase of the nuclear volume , Bcd is diluted in nuclei , which is partially offset by increase of incoming flux to the nuclei . Such dynamical balance among incoming/outgoing fluxes and dilution of Bcd should result in the gradual decrease of the Bcd concentration during the interphase period [11] . Since the growth of the nuclear size is not taken into account in the present model , decrease of Bcd concentration during the interphase period is not reproduced in simulation but the Bcd concentration gradually increases due to the active transport of Bcd into nuclei . Apart from such a difference , the overall features of the simulated results agree well with the observed data ( Figure 3D of Ref . 11 ) . The Bcd concentration is stably reproduced in the next nuclear cycle even after Bcd is lost during the mitotic period . This stability is quantified by comparing the peak values of the Bcd concentration in successive nuclear cycles . In Figure 12B , the peak values in successive nuclear cycles are shown to differ only about 10% , showing that the Bcd concentration profile is as stable as was observed in experiment ( Figure 3E of Ref . 11 ) . Such stable Bcd profile is realized because Bcd molecules released to cortical sites during the mitotic period are efficiently concentrated again into nuclei in the next nuclear cycle . The rapid transport of Bcd from cortex to nuclei is the necessary condition to prevent Bcd from escaping into core sites to achieve this efficient concentration . In Figure 12C , the simulated accumulation of Bcd in the cortical layer is compared with the experimental data ( Figure 6D of Ref . 11 ) . The factor Q ( m ) defined in Table 2 , which represents the gradual decrease of diffusion constant from cortex to core sites , was determined to reproduce this experimental plotting . See Method section for further explanation of Q ( m ) . In Figure S5 , the snapshots of the Bcd distribution are shown to visualize how Bcd is released in the mitotic period and is attracted into nuclei again in the interphase period . The Hb distribution is also stable in dynamically changing embryo . By calculating the distribution of the number of Hb molecules along a strip on the cylinder for t = 126–143 min in the interphase period of nuclear cycle 14 , we can see that the number of Hb molecules increases in a coherent manner along the AP axis as Bcd is being concentrated in nuclei and the Hb synthesis is promoted ( Figure S6A ) . The distribution profile of Hb is established soon after the Hb synthesis starts , so that the threshold position x1/2/L = ihalfΔx/L of the Hb distribution only slightly shifts during nuclear cycle 14 ( Figure S6B ) , which is consistent with the small shift of x1/2/L in the experimental data ( Figure 2C of Ref . 36 ) . The slope of the Hb distribution at x1/2 is kept steep during nuclear cycle 14 ( Figure S6C ) , which roughly agrees with the experimental data ( Figure 2D of Ref . 36 ) . The experimentally observed IO relation around x1/2 has been fitted by the Hill relation aswith the Hill coefficient n = 3–5 [12] , [35] , [36] , [38] . Results of the same fitting of the simulated data are shown in Figure 13A during nuclear cycle 11–14 . Through these nuclear cycles , n stays constant in interphase of different nuclear cycles in spite of the repeated mitoses between interphase periods . We should note that such stable Hb distribution should be realized in dynamically changing embryo only when the stable and precise patterning of Bcd distribution is established . Temporal change of fluctuation of the IO relation in nuclear sites near the threshold position x1/2 is shown in Figure 13B: As shown in Figure 11B , the fluctuation of the IO relation shows a peak at x1/2 . Plotted in Figure 13B is the time course of the peak height of the fluctuation of the IO relation . When Dh = 0 , the fluctuation sharply increases at the onset of interphase in each nuclear cycle because multiplicated nuclei are placed at positions away from Hb molecules inherited from the previous nuclear cycle . As the busting production takes place multiple times , the Hb molecules are accumulated around nuclei and the Hb concentration approaches the steady level due to the balance between the Hb production and degradation , but the remaining fluctuation is still larger than the fluctuation observed in experiment [12] . When Dh is finite with Dh = 0 . 1–0 . 5 µm2/s , on the other hand , the fluctuation is much suppressed and reaches the experimentally observed level [12] . Here , the important feature is that the fluctuation for nonzero Dh decreases by taking a long time from 50–35% in nuclear cycle 11 to 15–10% in nuclear cycle 14 . This long-time decrease is due to the slow diffusion of Hb . See also Video S1 to grasp the intuitive picture of how the slow diffusion of Hb averages the Hb distribution through multiple nuclear cycles to reach the homogeneous distribution in the surface layer of a slice of embryo . Though the larger Dh is more effective to reduce the fluctuation , too large Dh smears out the Hb distribution along the AP axis and flattens the slope of the profile at the boundary of hb activation as shown in Figure 6 for the one-dimensional model . To keep the sharpness of the Hb boundary , Dh has to be smaller than 0 . 3 µm2/s . As Dh is small , the self-averaging process of Hb should take time longer than the duration of a single nuclear cycle and multiple cycles are needed to reduce the fluctuation . In this way , the stable Bcd and Hb profiles lasting through multiple nuclear cycles is the basis to assure the effective reduction of fluctuation through multiple nuclear cycles .
One- and three-dimensional models of Drosophila embryo were developed and two major issues raised by the experimental visualization of embryonic development [11] , [12] were examined with stochastic simulations: One of the issues is on the mechanism to generate the profile of Bcd gradient . We showed that the stochastic processes of synthesis , diffusion and degradation of Bcd give rise to the Bcd distribution whose entire profile remains stable through multiple nuclear cycles . This stable profile is precise to exhibit only the small fluctuation within each embryo and is reproducible with small fluctuation among multiple embryos . The stable profile of Bcd is realized by the rapid transport of Bcd to nuclei . The other issue is on the readout of the Bcd gradient . Random diffusion and reception of Bcd molecules at the nanometer scale region of DNA induce the intense fluctuation in the hb gene activity . Our models showed that this fluctuation indeed dominates fluctuation in the hb expression , but the fluctuation in the Hb distribution is masked by the slow diffusion of Hb molecules: The slow diffusion of Hb through multiple nuclear cycles averages the Hb distribution without losing the sharp boundary of distribution at around the threshold position . In this way , self-averaging due to the slow diffusion of the output protein resolves the paradox of signal interpretation and enables the precise and reproducible readout of the gradient . Since the self-averaging process of Hb has to continue over multiple nuclear cycles to achieve the sufficient accuracy , the stable Bcd distribution over multiple nuclear cycles is a necessary condition for realizing the accurate Hb distribution . Thus , the coordinated diffusion of input and output molecules is the basis to generate the stable , precise , and reproducible patterning of both input and output molecules . Our prediction of the Hb self-averaging mechanism should be tested experimentally . Because mobility of the fused protein of Hb and the other protein domain should be affected by the characteristics of the added domain , diffusion constant of the Hb-fusion protein could be systematically changed by modulating the mass , size , and surface charge of the added domain . One possible example is to add a small tag such as histidine residues to eGFP , which can reduce velocity of the Hb-eGFP fusion protein [45] . Although such modulation may change the affinity of Hb to the hb enhancer to reduce effects of the feedback regulation of the hb expression , we should note that this feedback regulation has little effect on fluctuation in the Hb distribution as shown in Figure 7 and Figure S3 , so that the effects of the Hb self-averaging mechanism on fluctuation can be tested by systematically modulating the fused Hb proteins . Gregor et al . [12] also discussed the possible resolution of the paradox of signal interpretation through the self-averaging diffusion of Hb molecules . In their proposed mechanism , however , the positive feedback due to the binding of Hb protein to the hb enhancer was the necessary condition: They assumed that the hb genes in N≈50 neighboring nuclei are activated synchronously with the positive feedback interaction of the diffusing Hb and hb enhancers , thereby the effective size of the target DNA region is enlarged from a to aN , which leads to the modification of the criterion of Equation 1 to be . In the present simulations , however , we have not found an evidence of synchronized gene switching in multiple nuclei , so that this “self-averaging of input to the feedback loop” proposed in Ref . 12 is not necessary to resolve the paradox of signal interpretation . The paradox was resolved even in models without the positive feedback , which clearly shows that the simpler assumption of “self-averaging of output” is sufficient to resolve the paradox . In Drosophila embryos , molecules other than Bcd or Hb may also affect the hb gene activity [18] , [46] . The present results showed that assumption of the other input molecules is not necessary but the Bcd input and Hb output are sufficient to explain the stable , precise , and reproducible profiles of Bcd and Hb molecules at least in the anterior to middle regions of embryo . Several different measures have been used to evaluate the accuracy of the Hb distribution [6] , [9]–[12] , [17] , [36] , [47] , [48] . In this paper , we analyzed “preciseness” [12] of the Hb distribution by evaluating the smallness of fluctuation in each embryo , and “reproducibility” [6] , [10] , [12] by evaluating the smallness of fluctuation among multiple embryos . Other than these measures , “sharpness” [9] , [36] , and “robustness” [17] , [47] , [48] have been used to evaluate the accuracy . Our results showed that the sharp boundary of the Hb distribution is realized when Dh is within an optimal range: Dh should be smaller than 0 . 3 µm2/s to keep the average slope of the Hb boundary steep and Dh should be nonzero to suppress fluctuation around the average . In this way , the small fluctuation , or preciseness is a necessary condition for the sharpness . The present simulation showed that input of multiple kinds of transcriptional factors to hb is not necessary for the sharp Hb distribution . An experimental observation which is consistent with this result is the expression pattern of the reporter gene embedded into the zygotic genome [9]: The expressed pattern of the reporter gene which contains binding sites only for Bcd had a sharp boundary in embryos , which indicates that the other transcription factors than Bcd are not necessary for sharpness . This result is also consistent with the theoretical investigation: By solving the coupled equations of diffusion and reactions of Bialek and others [30]–[32] , we can see that the physical limit of Equation 1 is not changed when the hb enhancer receives multiple input molecules . Thus , the multiple input molecules do not by themselves resolve the paradox of signal interpretation but the mechanism of self-averaging diffusion is necessary for preciseness and hence for sharpness of the Hb distribution . Interactions among multiple genes , however , should be important for the Hb distribution to be robust against variations of conditions . Comparison between the observed and simulated data has suggested that the feedback regulations among gap genes are necessary for the robust Hb distribution [47] , [48] . This “canalization” mechanism , therefore , should be a necessary condition for robustness but our analyses suggest that the canalization mechanism by itself is not sufficient to assure preciseness and sharpness . We emphasize that the system described by the three-dimensional model is not in a steady state , so that the absolute values of concentrations at each position vary during nuclear cycles . Effects of non-steadiness were also focused on by Bergmann et al . [15] , [49] . Since the number of nuclei is small in the earlier stage of development , the fluctuation allowed to distinguish adjacent nuclei is about 50% in nuclear cycle 9 and 30% in nuclear cycle 11 by assuming an even distribution of nuclei during all nuclear cycles . Bergmann et al [49] suggested that the fate of each nucleus is determined as early as in nuclear cycle 9 . Such pre-steady state decoding should tolerate the large fluctuation because of the large spatial distance between nuclei . Our simulation data showed that the system is not in the steady state all through nuclear cycles as pointed out by Bergmann et al , but the fluctuation in the simulated IO relation was larger than 30% in nuclear cycle 11 when Dh = 0 . The pre-steady state decoding , therefore , does not resolve the paradox of signal interpretation by itself , but the slow diffusion of Hb is necessary for the precise decoding . The mechanism of diffusion of Bcd remains as an unsolved problem . From the experimental data of the fluorescence relaxation of Bcd-eGFP , diffusion constant was estimated to be Db<1 µm2/s [11] , but Db = 10–20 µm2/s was suggested by monitoring diffusion of inert molecules in embryo [10] . One possible explanation of this discrepancy is based on the difference of environment in these measurements: The fluorescence relaxation giving small Db was measured at the surface layer of embryo . Diffusion should be slow in such measurement because the complex and dense cytoskeletal structures in cortical layer trap the diffusing molecules [50] , [51] . Db≈10–20 µm2/s , on the other hand , was a result of diffusion through the unstructured core of embryo . The fluorescence relaxation in less structured unfertilized embryo , however , also showed Db<1 µm2/s , which casts doubts on such simple interpretation based on the difference of environment [11] . Another possible explanation is based on the notion of anomalous diffusion . Particles diffusing through medium of crowded obstacles should show the diffusion as for the short distance , and for the long distance with α<β [52] . If both and are fitted by and , the effective diffusion constant , , should become much smaller than , and hence it is possible to assume and . We can also expect that the cross-over distance between these two behaviors and other parameters such as and depend on the densities of cytoskeletal obstacles and take different values in cortical layer and core of embryos . Also possible is the mechanism that the cytoplasmic flow [14] , [53] enhances transport of molecules and increases the effective value of Db for the long spatio-temporal scale . Either type of the long-lived coherent flow [14] or the randomized flow should accelerate the movement of Bcd in embryo . Assumptions on values of diffusion constant in our models have not yet been checked from the microscopic viewpoint of the diffusion process , and further theoretical and experimental investigations are necessary to define diffusion constants on a sound basis . The recent experimental report suggested that bcd mRNA is not strictly localized at the anterior pole but forms a gradient by the mechanism of quasi-random active transport through a nonpolar microtublular network in cortex of the embryo [41] . If this is the case , Bcd can form the gradient by following its mRNA gradient with the relatively small diffusion constant , so that the assumption of the anomalous diffusion of Bcd may not be necessary . Test of this mechanism of dictation of the bcd mRNA gradient is possible with the present one- and three-dimensional models by changing the definition of sites which can synthesize Bcd and adjusting diffusion constants of Bcd . With such modifications , the mechanism of Bcd gradient formation will be altered in the model but we expect that the main results of the paper , such as the stable Bcd distribution through the rapid transportation of Bcd into nuclei and the precise and reproducible Hb distribution through resolution of the paradox , will not be changed because these results should be independent of whether the Bcd gradient is formed by following the bcd mRNA gradient or it is formed by the localized synthesis and diffusion . Another important problem is the scaling of distributions of developmental factors in embryo [54] . Several dipteran species have different size of embryos varying up to five fold , in which the distribution of Bcd scales with the length of embryo [10] , [17] , [34] , [55] , [56] . A possible mechanism of this scaling is the domination of degradation of Bcd at nuclei to determine the Bcd profile [10] , [11] . Because the positional distribution of nuclei scales with the length of embryo , one might consider that this assumption leads to the scaling of Bcd distribution . We examined this hypothesis with our three-dimensional model , but the results indicated that the Bcd distribution does not scale even when nuclei distribution scales ( data not shown ) . To solve this problem , the assumption of the larger rate of trapping to the nuclei might be needed [57] , or the scaling of transport or diffusion mechanism of Bcd or bcd mRNA should be necessary . Examination of these hypotheses is left for future investigations .
|
For developing embryos , the precise , position-specific regulation of molecular processes is of fatal importance . As the mechanism of such regulation , widely accepted has been the notion of the intraembryonic distribution of regulatory molecules called “morphogens” . One of the best-studied morphogens is Bicoid in the early developmental stage of the Drosophila embryo . Synthesized around the anterior pole of the embryo , Bicoid forms an exponential gradient of concentration to initiate expression of a target gene , hunchback , in nuclei at the periphery of the embryo . This invariably forms a concentration boundary of the product protein Hunchback at around 49% embryo length . Remarkably , the embryo-embryo variability in the boundary position is less than 5% . Reactions in embryos , however , should be intrinsically noisy because the number of molecules involved is small , and those reactions are governed by randomly diffusing molecules . The mechanisms to generate the invariable Hunchback distribution by filtering the intense noise remain mysterious , and here we construct models to shed light on this problem . Stochastic simulations show that the slow diffusion of Hunchback averages out the intense noise , so that the coordinated rates of diffusion and transport of input Bicoid and output Hunchback play decisive roles in suppressing fluctuations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion"
] |
[
"computational",
"biology/systems",
"biology",
"computational",
"biology/transcriptional",
"regulation",
"cell",
"biology/morphogenesis",
"and",
"cell",
"biology",
"computational",
"biology/molecular",
"genetics",
"biophysics/theory",
"and",
"simulation",
"cell",
"biology/developmental",
"molecular",
"mechanisms",
"biophysics/transcription",
"and",
"translation",
"cell",
"biology/gene",
"expression"
] |
2009
|
Stable, Precise, and Reproducible Patterning of Bicoid and Hunchback Molecules in the Early Drosophila Embryo
|
Phylodynamic models are widely used in infectious disease epidemiology to infer the dynamics and structure of pathogen populations . However , these models generally assume that individual hosts contact one another at random , ignoring the fact that many pathogens spread through highly structured contact networks . We present a new framework for phylodynamics on local contact networks based on pairwise epidemiological models that track the status of pairs of nodes in the network rather than just individuals . Shifting our focus from individuals to pairs leads naturally to coalescent models that describe how lineages move through networks and the rate at which lineages coalesce . These pairwise coalescent models not only consider how network structure directly shapes pathogen phylogenies , but also how the relationship between phylogenies and contact networks changes depending on epidemic dynamics and the fraction of infected hosts sampled . By considering pathogen phylogenies in a probabilistic framework , these coalescent models can also be used to estimate the statistical properties of contact networks directly from phylogenies using likelihood-based inference . We use this framework to explore how much information phylogenies retain about the underlying structure of contact networks and to infer the structure of a sexual contact network underlying a large HIV-1 sub-epidemic in Switzerland .
From the viewpoint of an infectious pathogen , host populations are highly structured by the physical contacts necessary for disease transmission to occur . For pathogens whose transmission does not require intimate or sustained physical contact , random mixing models assuming contacts form instantaneously between individuals may offer a reasonable approximation to the true dynamics of person-to-person contact [1–3] . But for pathogens like sexually-transmitted infections ( STIs ) , transmission requires contacts that are generally more limited in number , less transient in nature , and form non-randomly based on individual behavior—resulting in host populations that are highly structured locally at the level of individuals [4–7] . Even for non-STIs , a limited number or clustering among contacts may constrain transmission; potentially explaining why newly emerging infections like SARS and Ebola give rise to large outbreaks in some social settings but not others [8–10] . It is therefore often more reasonable to view communities as networks of individuals connected by edges that represent the physical contacts through which transmission can occur . Through the study of theoretical network models , epidemiologists now understand that contact network structure has a profound influence on epidemic dynamics and whether or not control strategies will be effective [8 , 11–14] . Yet studying the structure of contact networks empirically through methods such as contact tracing is difficult and costly , meaning we often know little about the structure of contact networks underlying real-world epidemics [15 , 16] . New hope for the empirical study of contact networks has emerged in recent years from the widespread availability of pathogen molecular sequence data . In molecular epidemiology , sequence data is already commonly used to link individuals into probable transmission pairs or clusters based on the phylogenetic distances between their pathogens . While such approaches do not directly reveal the structure of contact networks , they can reveal paths in the contact network through which the pathogen spread and provide a useful heuristic for assessing how well connected networks are within and between different subpopulations or risk-groups [17–19] . Other methods in molecular epidemiology attempt to reconstruct the full details of the underlying transmission tree , the directed graph showing exactly who infected whom in an outbreak [20–25] . Essentially though , all current methods for inferring linkage and transmission trees take a bottom-up approach—they attempt to reconstruct routes of transmission by linking sampled individuals based on their phylogenetic distance . While this can be a powerful approach for studying densely sampled outbreaks where most infected individuals are sampled , bottom-up approaches may provide misleading results when applied to sparsely sampled epidemics . In this case , two infected individuals may have pathogens that are most closely related to one another phylogenetically but an unknown number of intervening infections might separate them in the true transmission tree . Thus the phylogenetic proximity of individuals may only weakly correlate with their proximity in the transmission tree , making it very difficult to reconstruct the detailed transmission history of who infected whom . While it may not be possible to reconstruct the detailed structure of transmission networks from sparsely sampled data , it may still be possible to infer large-scale properties of contact networks . By simulating the phylogenetic history of pathogens spreading through networks , recent studies have shown that network properties can exert a strong influence on the structure of phylogenetic trees [26–28] . For example , increasing levels of contact heterogeneity—variation in the number of contacts individuals form—can result in increasingly asymmetric or imbalanced trees and shift the distribution of coalescent ( i . e . branching ) events earlier towards the beginning of an epidemic [26 , 27] . However , statistical measures of tree topology like imbalance may only weakly correlate with network statistics like contact heterogeneity , and may also be highly dependent on how samples are collected [28] . Moreover , in addition to network structure , population dynamics also strongly shape phylogenies and therefore potentially confound inferences of network structure drawn from phylogenies [28] . For example , clustering of samples together in phylogenetic trees has previously been assumed to indicate clustering of individuals in the underlying contact network , but phylogenetic clustering can arise naturally in epidemics even when no measurable degree of clustering exists in a population [29] . Taken together then , previous work suggests that contact network structure can shape pathogen phylogenies , but we do not yet know how to properly extract this information from trees . In this paper , we present a new theoretical framework for relating pathogen phylogenies to contact networks using phylodynamic modeling . Our approach is quite different from bottom-up approaches in that it does not attempt to reconstruct the details of person-to-person transmission . Rather , we start with a random graph model [30] that captures the important statistical properties of real-world networks . We then use pairwise epidemic models [31–33] to capture the population dynamics of an epidemic on a network with the statistical properties specified by the random graph model . In addition to tracking the infection status of individuals , these pairwise models track the status of pairs of individuals and thereby correlations in the infection status of neighboring individuals , such as the depletion of susceptible hosts around infected individuals . Analogously , by shifting our focus from the level of individuals to the level of pairs , we derive a relatively simple coalescent model that captures a pathogen’s phylogenetic history as a backwards-time dynamical process on a network . The pairwise coalescent model naturally takes into account incomplete sampling and how network structure and epidemic dynamics interact to shape pathogen phylogenies . By considering phylogenies in a probabilistic framework , the pairwise coalescent model also allows us to compute the likelihood of a given phylogeny evolving on a network with defined statistical properties , and therefore to estimate the structure of networks from phylogenies using likelihood-based inference . How local contact network structure shapes pathogen phylogenies has received some attention in recent years [26–28] , but has not been comprehensively studied . After deriving the pairwise coalescent model , we therefore begin by using simulations to explore how network properties such as overall connectivity , clustering , contact heterogeneity and assortativity shape phylogenies . Using these simulations , we demonstrate that the pairwise coalescent model captures how these network properties shape phylogenies in terms of coalescent times , how lineages move through a network , and overall tree topology . We then go on to show that the model can be used to accurately estimate network properties from phylogenies , although how precisely depends strongly on sampling effort . Finally , we have implemented the model in BEAST 2 [34] as a package called PairTree , which we use to estimate the structure of a contact network underlying a large HIV sub-epidemic among men-who-have-sex-with-men ( MSM ) in Switzerland .
In network epidemiology , random graph models are often used to model the large-scale statistical properties of networks while treating the fine-scale details of who is connected to whom as random . Random graph models can therefore be thought of as a probability distribution on graphs constrained to take on certain statistical properties . Here , we use the configuration model [35] and extensions thereof to model network structure and generate random graphs parameterized to vary in overall connectivity , clustering , contact heterogeneity and assortative mixing . The second component of our modeling framework consists of epidemiological models that describe the dynamics of a pathogen spreading through a network with statistical properties specified by a random graph model . As in standard SIR-type epidemiological models , we track the infection status of each host node as susceptible or infected , along with an optional recovered class . We use the notation [Sk] and [Ik] to denote the number of degree k susceptible and infected individuals . [SkIl] denotes the the number of pairs in the network connecting susceptible individuals with degree k to infected individuals with degree l . At the level of individuals , the epidemic dynamics are described by the following differential equations: d [ S k ] d t = - τ ∑ l [ S k I l ] + δ 0 , 1 ν [ I k ] d [ I k ] d t = τ ∑ l [ S k I l ] - ν [ I k ] ( 5 ) Here , τ is the per-contact rate at which infected individuals transmit to their neighbors and ν is the removal or recovery rate . The dummy variable δ0 , 1 is set to either 1 or 0 depending on if recovered individuals become susceptible again ( the SIS model ) or are immunized ( the SIR model ) . As seen from Eq ( 5 ) , the transmission dynamics depend on the [SkIl] terms and thus how individuals are connected into pairs or partnerships . We therefore need to track the dynamics at the level of pairs , which in turn depends on the number of triples such as [SkSl Im] where m is the degree of the third node: d [ S k S l ] d t = - τ ∑ m [ S k S l I m ] + [ I m S k S l ] + δ 0 , 1 ν ( [ S k I l ] + [ I k S l ] ) d [ S k I l ] d t = τ ∑ m ( [ S k S l I m ] - [ I m S k I l ] ) - τ [ S k I l ] - ν [ S k I l ] + δ 0 , 1 ν [ I k I l ] d [ I k I l ] d t = τ ∑ m ( [ I k S l I m ] + [ I m S k I l ] ) + τ ( [ S k I l ] + [ I k S l ] ) - 2 ν [ I k I l ] ( 6 ) These are the pairwise network equations introduced by [31] and extended to heterogenous contact networks by [33] . By tracking the status of pairs rather than just individuals , the pairwise equations take into account local correlations that build up over time between the infection status of neighboring nodes; hence their other common name , correlation equations [32] . These local correlations arise because a node’s infection status depends strongly on the status of its neighbors . For example , early on in an epidemic positive correlations develop between infected individuals , reflecting the fact that infected individuals are likely to be surrounded by other infected individuals who either infected them or became infected by them . Because these correlations can have a strong impact on epidemic dynamics , such as through the local depletion of susceptible nodes surrounding infected nodes , tracking these correlations allows pairwise models to more accurately describe epidemic dynamics on networks . The initial conditions for the pairwise epidemic model depend on the degree distribution dk and edge degree distribution ekl of the network and are described in S1 Text . While the dynamics at the level of pairs depends on the number of triples , which in turn depends on even higher-order configurations , previous work has shown that moment closure methods can be used to approximate the number of triples based on the number of pairs without much loss of accuracy [33 , 39] . We thus “close” the system at the level of pairs by approximating each triple of arbitrary type [ABC] as: [ A B C ] = l - 1 l [ A B ] [ B C ] [ B ] ( 1 - ϕ ) + N l [ A C ] [ A ] [ C ] ϕ , ( 7 ) where l is the degree of the central node in state B . By taking into account the clustering coefficient ϕ , this moment closure takes into account additional state correlations that can arise between three nodes when there is appreciable clustering in the network [31 , 32] . The basic reproductive number R0 , as usually defined , is the average number of secondary cases resulting from a single infectious individual in an otherwise susceptible population . As shown in [33] , for the pairwise model on a heterogenous network , R 0 = τ ( λ m a x - 1 ) ν , ( 8 ) where λmax is the dominant eigenvalue of the next generation matrix M , which has elements M k l = ( l - 1 ) [ N k N l ] l [ N l ] . ( 9 ) Here , Nk and Nl represent the total number of individuals in the network with degree k and l . The third and novel component of our modeling framework are coalescent models that allow us to probabilistically relate the phylogenetic history of a pathogen back to the dynamics of an epidemic on a network . In essence , these coalescent models provide a probability distribution over trees , and therefore allow us to compute the likelihood of a given phylogeny having evolved on a random graph with known statistical properties . While coalescent theory has previously been extended to accommodate the nonlinear transmission dynamics of infectious pathogens [40–43] , these coalescent models assume random mixing , at least within discrete subpopulations , and therefore neglect local contact network structure . Below , we extend the structured coalescent framework of [43] to include local contact network structure by shifting our focus from the level of individuals to pairs of hosts in the network . The likelihood of a time-scaled phylogeny T under a structured coalescent model with parameters θ has the general form: L ( T | θ ) = ∏ p = 1 P - 1 λ i j ( t p ) exp -∫s=tps=tp+1∑ii∈A ( s ) ∑j>ij∈A ( s ) λij ( s ) ds . ( 10 ) For a tree containing P samples , the total likelihood is the product of the likelihood of each of the P − 1 coalescent events and the waiting times between events . The likelihood of each coalescent event depends on the rate λij ( tp ) at which lineages i and j coalesce at time tp . These rates may depend on the location of lineages i and j , and thus this formulation of the coalescent likelihood accommodates population structure . The likelihood of the waiting time between coalescent events is given by the exponential term in Eq ( 10 ) . Here , the sums are over all lineages A ( s ) present in the phylogeny at time s , which is allowed to change within coalescent intervals due to sampling . The pairwise coalescent rates λij are centrally important to our model as they are required to compute the likelihood in Eq ( 10 ) and provide the main link between the epidemic dynamics and the coalescent process . To derive these rates , we begin by making the simplifying assumption common in phylodynamics that only a single pathogen lineage resides in each infected host . While this assumption ignores within-host pathogen diversity , it dramatically simplifies the relationship between transmission events and coalescent events in the pathogen phylogeny: each coalescent event in the phylogeny will represent a transmission event on the network . Below , we use this relationship to derive the pairwise coalescent rates λij for pairs of lineages . To summarize , our approach uses a random graph model to describe the statistical properties of a network including its degree distribution . These statistical properties are then used to initialize the state variables in the pairwise epidemic model—including the degree distribution of susceptible and infected individuals in the network . The ODEs for state variables like the number of [SkIl] pairs given in Eqs ( 5 ) and ( 6 ) are then solved forward in time . Given these forward-time dynamics , the lineage state probabilities can be solved backwards in time along each lineage in the phylogeny according to Eq ( 17 ) . With both the epidemic state variables and lineage state probabilities at hand , the pairwise coalescent rates can be computed according to Eq ( 14 ) . Finally , with these coalescent rates the likelihood of a phylogeny can be calculated using Eq ( 10 ) , allowing us to infer network parameters in either a maximum likelihood ( ML ) or Bayesian framework . For ML inference , we use a numerical optimization routine ( fminsearch in Matlab ) to find the parameter values that maximize the likelihood of a given phylogeny . For Bayesian inference , we use a MCMC approach to sample from the posterior distribution . For the latter , the pairwise coalescent model was implemented in BEAST 2 . Details on how the MCMC analysis was performed are provided in the input XML files provided along with the source code at https://github . com/davidrasm/PairTree .
To see how key network properties shape pathogen phylogenies and how well the pairwise coalescent model captures their effects , we generated networks under random graph models parameterized to obtain networks with known statistical properties . On top of these networks , we simulated the spread of an epidemic using individual-based stochastic ( IBS ) simulations that tracked the ancestry of each pathogen lineage forward in time so that a true phylogeny was obtained from each simulation ( see S1 Text ) . We then compared the epidemic dynamics and phylogenies simulated under the IBS model to those expected under the pairwise epidemic and coalescent models . As expected from earlier work [14 , 39 , 44] , the pairwise epidemic model provides an excellent deterministic approximation to the mean dynamics observed in IBS simulations across a wide range of random networks , whereas random mixing models generally do not ( Fig 1 ) . Likewise , the pairwise coalescent model does an excellent job of capturing the coalescent process on these networks in terms of the temporal distribution of coalescent events over the epidemic ( Fig 2 , blue ) . In contrast , the coalescent distributions expected under a random mixing coalescent model provide a reasonable approximation on some networks but not others ( Fig 2 , red ) . For example , on poorly connected and highly clustered networks , the expected distribution of coalescent times under random mixing deviates widely from the IBS simulations . This is the case even if we condition the random mixing model on the more accurate population trajectories predicted by the pairwise epidemic model ( Fig 2 , green ) . On better connected networks and on networks with more contact heterogeneity , the random mixing model does almost as well as the pairwise coalescent model . In the S1 Text , we additionally explore when the pairwise approximation fails due to the presence of higher-order network structure ( see S2 Fig ) . In addition to coalescent time distributions , contact network structure can shape the topology of phylogenies . In particular , pathogens residing in well connected hosts may have the opportunity to infect more hosts and therefore leave more descendent lineages , causing trees to become increasingly asymmetric or imbalanced as the amount of contact heterogeneity increases in a population [26–28 , 45] . We therefore simulated trees on random networks with different levels of contact heterogeneity using IBS simulations and under the pairwise coalescent model using backward-time simulations in order to see if the coalescent model can capture the effects of contact heterogeneity on tree imbalance . Trees were compared using three of the most widely used measures of tree imbalance [46]: Colless’s index , Sackin’s index and the number of cherries . How these statistics were computed and normalized for trees of different sizes is described in the S1 Text . Trees generated by IBS simulations and under the pairwise coalescent model both grow increasingly imbalanced with increasing contact heterogeneity ( Fig 3 ) . Colless’ and Sackin’s index both increase with greater contact heterogeneity , while the number of cherries decreases as expected for more imbalanced trees . However , IBS trees are more imbalanced overall than coalescent trees and grow disproportionally more so with increasing contact heterogeneity . This discrepancy may be due to additional network structure above the level of pairs not accounted for in the pairwise models—leading to additional variability in transmission potential for lineages in different parts of a network . Thus , it appears that the pairwise coalescent can partially capture the effects of local contact structure on tree topology , although the effect of network structure appears to be stronger in trees evolving on actual networks . The pairwise coalescent model allows us to compute the likelihood of a given phylogeny being generated by an epidemic on a random graph with defined statistical properties . It is therefore possible to estimate the statistical properties of a network directly from a phylogeny . However , phylogenies may retain little information about contact network structure , especially if the epidemic is sparsely sampled . To explore the information content of phylogenies regarding network structure , we simulated epidemics on random networks with known statistical properties . A variable fraction of infected nodes was then sampled upon removal to obtain phylogenies with sampling fractions ρ of 10 , 25 , 50 and 100% . The pairwise coalescent model was then used to construct likelihood profiles for each parameter controlling local network structure while all other epidemiological parameters were fixed at their true values . At sampling fractions at or below 10% , except for overall connectivity the simulated phylogenies contain little or no information about local network structure , as seen from the essentially flat likelihood profiles ( Fig 4 ) . At sampling fractions ≥ 25% , the likelihood profiles begin to show significant curvature for clustering and contact heterogeneity , and with sampling fractions ≥ 50% the likelihood profiles are sharply curved enough that these parameters can be estimated rather precisely with narrow 95% confidence intervals . Assortativity appears more difficult to infer from phylogenies , even if the true degree of sampled nodes is provided . Although the likelihood profiles for r do show some curvature at sampling fractions ≥ 50% , the credible intervals remain relatively wide even with complete sampling . To check for potential biases in our estimates of network parameters , we simulated 100 additional phylogenies under a fixed value of each parameter using forward-time IBS simulations . We then obtained a maximum likelihood estimate ( MLE ) of the corresponding parameter . The MLEs appear centered around the true parameter values with little to no detectable bias ( Fig 5a–5d ) . Next , we simulated trees under a wider range of parameter values for each network property to check how well our estimator performs under different model parameterizations . Overall , parameter estimates appear well-calibrated with a high correlation between the true and MLE values ( Fig 5e–5h ) . While the coverage of our confidence intervals falls below the desired 95% level , we believe the coverage achieved is very reasonable given that the epidemic dynamics in the stochastic simulations can diverge considerably from what is expected under the pairwise model . In S1 Text , we in fact show that most of the estimation error can be attributed to stochastic variation in the epidemic dynamics ( see S3 Fig ) . We performed a phylodynamic analysis of a HIV-1 subtype B epidemic among men-who-have-sex-with-men ( MSM ) in Switzerland using the pairwise coalescent model to see if we could estimate the statistical properties of a real-world sexual contact network . HIV pol sequences from infected individuals were obtained from patients enrolled in the Swiss HIV Cohort Study [47–49] . To minimize the effects of spatial structure , we focus on a single large sub-epidemic identified as primarily occurring in the Zürich region in a preliminary phylogenetic analysis ( see S1 Text and S4 Fig ) . A time-calibrated phylogeny containing 200 sequences revealed this sub-epidemic to be quite genetically diverse with many older lineages originating in the early 1980’s ( Fig 6a ) . We fit a SIR-type pairwise epidemic model to this sub-epidemic while simultaneously inferring the tree from the sequence data in BEAST 2 . We assumed a discretized gamma distribution for the degree distribution dk , which allowed us to independently estimate the mean μk and variance σ k 2 of the network’s underlying degree distribution . The posterior estimates of μk and σ k 2 indicate that the network was not especially well-connected ( median μk = 2 . 20 ) but heterogenous in degree ( median σ k 2 = 4 . 04 ) ( Fig 6b and 6c ) . The basic reproductive number was estimated to be between 1 . 0 and 2 . 5 , although unlike for the network parameters the posterior density of R0 only diverged slightly from the prior ( Fig 6d ) . R0 values estimated under the pairwise coalescent model were however significantly lower than the values estimated under the random mixing model , even though the same prior on R0 was used for both models . Overall our phylodynamic analysis suggests that this particular sub-epidemic spread rapidly by way of a few highly connected individuals . This is supported by the inferred degree distribution of the network and can be seen from the expected degree of lineages computed from the inferred ancestral degree distribution of each lineage over time ( Fig 6a ) . Most coalescent ( i . e . transmission ) events early in the epidemic are attributable to lineages residing in high degree individuals . Later , towards the beginning of the 2000’s , a few clusters in the tree begin to grow again through new transmission events along lineages with higher than average degree , which corresponds in time to the resurgence of HIV among MSM in Switzerland [48 , 50] .
Recent work has suggested that the structure of local contact networks can shape pathogen phylogenies [26–28] . Yet it remains unclear how much information pathogen phylogenies retain about the networks through which they spread and how to best extract information about network structure from trees . As a step towards addressing these questions , we sought a simple theoretical framework to explore the relationship between contact networks , epidemic dynamics , and phylogenies . Starting with random graph and pairwise epidemic models , we derived a fairly simple coalescent model that includes local network structure by using pair approximations . By treating the coalescent process as a backwards-time dynamical process on a network , our pairwise coalescent model allows us to capture the phylogenetic history of a pathogen in terms of how lineages move through a network and the rates at which they coalesce . As we have shown , our phylodynamic modeling framework provides a very good approximation to the coalescent process on random networks and can recapitulate the major features of pathogen phylogenies simulated on different types of random graphs . Using the pairwise coalescent model and individual-based stochastic simulations as guides , we reexamined how contact network structure shapes pathogen phylogenies . We found that local contact network structure can have a strong impact on the the coalescent process in terms of the timing of coalescent events . Network properties like overall connectivity and contact heterogeneity that increase the epidemic growth rate concentrate coalescent events towards the beginning of an epidemic , while properties like clustering that slow epidemic growth broaden the distribution of coalescent events over the epidemic . On the other hand , properties like assortativity that were observed to have no strong effect on epidemic dynamics likewise have little influence on the timing of coalescent events ( although we may have observed a larger effect if we had considered more extreme forms of assortative mixing [51–53] ) . This suggests that local contact network structure primarily shapes the coalescent process indirectly through the network’s influence on epidemic dynamics , particularly the timing of transmission events . Because the pairwise coalescent model can be used for likelihood-based inference , it also offers a means of exploring how much information phylogenies contain about contact network structure . Using simulated phylogenies , our ability to infer network properties was highly dependent on the fraction of sampled individuals . While we could estimate network properties that strongly regulate epidemic dynamics such as overall connectivity even at sampling fractions as low as 10–25% , other properties that do not strongly regulate epidemic dynamics like assortativity proved difficult to precisely estimate even with complete sampling . This observation suggests that for the parameters that can be estimated at low sampling fractions , we may largely be inferring the structure of networks not from any direct signal of network structure in the tree itself , but from the indirect effect of network structure on the epidemic dynamics reflected in the timing of the coalescent events in the tree . Thus , without additional information beyond a phylogeny , network parameters may essentially be statistically unidentifiable from other epidemiological parameters like transmission rates that control epidemic dynamics . Although it appears difficult to estimate some network properties from phylogenies , we were able to obtain informative estimates of the degree distribution of a sexual contact network underlying a large HIV sub-epidemic in Switzerland . While we were likely helped by the high fraction of HIV infected individuals sampled in Switzerland ( at least 34–38% ) compared with other countries and the relatively informative priors we placed on the model’s epidemiological parameters , this demonstrates that it is at least technically possible to estimate the structure of real-world contact networks from phylogenies . Our phylodynamic estimates indicated that the Swiss MSM network was not particularly well-connected with a rather low mean degree , but that the number of contacts per individual was highly variable . While a high degree of contact heterogeneity is consistent with other empirical studies of sexual contact networks [4 , 6 , 54] , such low overall connectivity may be surprising in a high-risk population such as MSM . However , our estimates were based on a random graph model assuming a completely static network and therefore may better reflect the momentary degree distribution of the network in terms of the number of concurrent or nearly concurrent contacts individuals have rather than their total number of partners over time , which may grow substantially larger depending on the time they remain sexually active . As in mathematical epidemiology , phylodynamics has made conceptual progress by developing simple models that largely assume random mixing [40 , 55 , 56] . It may now be worthwhile to consider when including network structure is likely to be important in phylodynamics . From earlier work in network epidemiology , it is well known that random mixing models can fail to provide an accurate description of epidemic dynamics because they ignore how local structure promotes or hinders transmission as well as how this local structure evolves over the course of an epidemic [14 , 31 , 57] . The timing of transmission events as regulated by local interactions on the network also appears to determine how well random mixing models can approximate the coalescent process on networks . On weakly connected or highly clustered networks where local interactions strongly limit transmission due to saturation effects , random mixing models overshoot the true transmission rate and therefore also the expected coalescent rate . In better connected networks , the effect of these local interactions on transmission is minimized by well-connected nodes and random mixing models can perform quite well [14] . The effect of local contact network structure on the coalescent process can therefore probably be safely ignored in highly connected networks , but may be important to consider in less well-connected networks . While network structure can also shape the topological characteristics of phylogenies such as tree imbalance , these effects appear to be quite weak [28] , unless the distribution of contacts is extreme , as in the case of scale-free networks with power law degree distributions [27] . For multi-strain pathogen systems , network structure can have more complex effects on evolutionary dynamics depending on how different pathogens interact; ranging from a higher probability of invading strains being competitively excluded [58 , 59] to facilitating the spread of synergistic co-infections [60] . The importance of multi-strain interactions on networks remains to be explored in phylodynamics . Considering network structure may also be important when estimating and interpreting key epidemiological parameters . Of perhaps most interest to epidemiologists is R0 , which is a function of both the transmissibility of a disease and contact network structure [8] . Thus , even if two pathogens are equally transmissible per contact , they can have substantially different R0 values based on contact patterns . Contact heterogeneity can for instance dramatically increase R0 , while clustering can reduce R0 [61 , 62] . But network structure can have an even more dramatic impact on epidemic dynamics , especially initial growth rates . For example , epidemics can spread very rapidly though heterogenous contact networks via a few highly connected individuals . In this case , R0 would likely be overestimated under a random mixing model because fast initial growth can only be explained under such a model by uniformly high transmission rates in the population; whereas in fact the highly connected individuals infected early in the epidemic do not represent the average connectedness of the population . This most likely explains why we estimated R0 for the Swiss HIV sub-epidemic to be more than twice as high under a random mixing model than under a model accounting for contact heterogeneity . Phylodynamic inference using random mixing models may be especially prone to large estimation errors in such settings because most phylogenetic information is concentrated in the early stages of an epidemic when most coalescent events occur just as network structure begins to shape epidemic dynamics . While we strove for simplicity , the true complexity of real-world contact networks does highlight some deficiencies in the pairwise models . First , while we only considered perfectly static random graph models , real-world networks temporally evolve as new contacts form and dissolve . Pairwise epidemic models that allow for dynamic partner exchange have been proposed [63 , 64] , and in theory could be merged with our pairwise coalescent model to explore contact durations that are intermediate between the infinitesimal nature assumed by random mixing models and the permanent nature assumed by static models , as in [65] . Second , the deterministic pairwise models ignore the often considerable stochastic variability of epidemics on networks , which we found to be a major source of estimation error . Particle filtering approaches similar to those in [66] could be adapted to the pairwise coalescent model , although it is not straightforward to simply add stochasticity into pairwise models [67 , 68] . Finally , the random graph models we employed here only consider local structure at the level of pairs in the network . Higher-order structure that subdivides networks into different communities most likely also plays a very strong role in shaping pathogen phylogenies . Developing methods that can quantify connectivity within and between communities while accounting for epidemic dynamics and incomplete sampling such as our approach does on local networks remains a challenging but highly important area of future research .
|
Phylodynamic models relate the branching pattern of a pathogen’s phylogenetic tree to the tree-like growth of an epidemic as it spreads through a host population . Such models are increasingly used to learn about the epidemiology of different pathogens . We extend current models to consider the structure of host contact networks—the web of physical interactions through which pathogens spread . By considering how local interactions among hosts shape the phylogeny of a pathogen , our models offer a “pathogen’s eye view” of these networks . Our models also provide a statistical framework that can be used to infer network structure directly from phylogenies , which we use to estimate the properties of a sexual contact network in Switzerland from a HIV phylogeny .
|
[
"Abstract",
"Introduction",
"Models",
"and",
"methods",
"Results",
"Discussion"
] |
[
"taxonomy",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"infectious",
"disease",
"epidemiology",
"pathogens",
"geographical",
"locations",
"epidemiological",
"methods",
"and",
"statistics",
"simulation",
"and",
"modeling",
"phylogenetics",
"data",
"management",
"phylogenetic",
"analysis",
"research",
"and",
"analysis",
"methods",
"europe",
"hiv",
"epidemiology",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"epidemiology",
"evolutionary",
"systematics",
"switzerland",
"epidemiological",
"statistics",
"people",
"and",
"places",
"biology",
"and",
"life",
"sciences",
"evolutionary",
"biology"
] |
2017
|
Phylodynamics on local sexual contact networks
|
The evolution of cooperation described in terms of simple two-person interactions has received considerable attention in recent years , where several key results were obtained . Among those , it is now well established that the web of social interaction networks promotes the emergence of cooperation when modeled in terms of symmetric two-person games . Up until now , however , the impacts of the heterogeneity of social interactions into the emergence of cooperation have not been fully explored , as other aspects remain to be investigated . Here we carry out a study employing the simplest example of a prisoner's dilemma game in which the benefits collected by the participants may be proportional to the costs expended . We show that the heterogeneous nature of the social network naturally induces a symmetry breaking of the game , as contributions made by cooperators may become contingent on the social context in which the individual is embedded . A new , numerical , mean-field analysis reveals that prisoner's dilemmas on networks no longer constitute a defector dominance dilemma—instead , individuals engage effectively in a general coordination game . We find that the symmetry breaking induced by population structure profoundly affects the evolutionary dynamics of cooperation , dramatically enhancing the feasibility of cooperators: cooperation blooms when each cooperator contributes the same cost , equally shared among the plethora of games in which she participates . This work provides clear evidence that , while individual rational reasoning may hinder cooperative actions , the intricate nature of social interactions may effectively transform a local dilemma of cooperation into a global coordination problem .
Portuguese is no exception: Like any other language , it has many proverbs and popular sayings . One of them states something like: I have already contributed to that charity [1] , concerning originally situations in which individuals are faced with the decision of offering ( or not ) a contribution to a common venture , the expression above meaning “no” . Interestingly , the amount given is never stated . It turns out that , quite often , we are confronted with situations in which the act of giving is more important than the amount given . Let us keep with a charity event , in which some celebrities are invited to participate . Typically their appearance is given maximal audience , and they are shown contributing a seemingly large amount of money to the charity's cause . This offer is aimed at stimulating the contribution of many to the same charity , and indeed this mechanism of “celebrity participation in charities” is common , and presumably effective . But what is the relevance of the amount contributed by the celebrity ? It is certainly impressive to many , despite being , most likely , a small contribution , both in face of the celebrity's wealth and also in what concerns the overall amount accumulated . But it does induce , hopefully , a large number of ( much smaller ) contributions from anonymous ( non-celebrities , the overwhelming majority ) charity participants , who feel compelled to contribute given the fact that their role model ( the celebrity ) contributed . In other words , the majority copies ( imitates ) the act of giving , but certainly not the amount given . Nowadays , web-signed petitions are also examples of collective decisions which , often , benefit from the fact that some well-known people adhere to the petition's cause . Besides those who are fully aware and agree with the cause , there are also those who sign the petition simply because they admire someone who has signed the petition , again copying the attitude . Many other examples from real life could be provided along similar lines , from trivia , to fads , to stock markets , to Humanitarian causes up to the salvation of planet Earth [2]–[4] . From a theoretical perspective , many of these situations provide beautiful examples of public goods games [5] , [6] ( PGG ) which are often hard to dissociate from reputation building , social norms and moral principles [7]–[11] . This intricate interplay reflects the many-body nature and multi-level complexity of the interactions among the “social atoms” [12] . The simplest PGG involves two persons . Both have the opportunity to contribute a cost c to a common pool . A Cooperator ( C ) is one who contributes; otherwise she is a Defector ( D ) . The total amount is multiplied by an enhancement factor F and equally shared between the two participants . Hence , player i ( i = 1 , 2 ) using strategy si ( si = 1 if C , 0 if D ) gets a payoff from this game , leading to the following payoff matrix ( 1 ) For Ds dominate unconditionally . For F = 2 no strategy is favored in well mixed populations ( neutral drift ) ; yet , for , it is better to play C despite the fact that , in a mixed pair , a D collects a higher payoff than a C . For the game is an example of the famous symmetric one-shot two-person prisoner's dilemma [13] , on which many central results have been obtained over the years , in particular in the context of evolutionary game theory [14] , [15]: In 1992 [16] it has been explicitly shown that population structure matters , despite its importance being recognized already by Darwin , albeit in the form of Group selection [17] , [18] . It clearly makes a difference whether everybody is equally likely to interact with anybody else in the population or not ( see also [19] ) . In 2004 we learnt that evolutionary game theory in finite populations may behave very differently from that on infinite populations [20] , even in the absence of any population structure , Evolutionarily Stable Strategies ( ESS ) becoming population size dependent . In 2005 we learnt that heterogeneous population structures play an important role in the evolution of cooperation under the prisoner's and other social dilemmas [21] , [22] , a result which spawned a number of investigations [23]–[29] ( see also Szabó and Fáth for a recent review [30] ) . In 2006 a mathematical condition was obtained for Cs to become advantageous on populations structured along the links of homogeneous networks [31] , subsequently confirmed making use of inclusive fitness methods [32] for a limited subset of game payoff matrices . This result , valid in the limit of weak selection , has also unraveled an important feature of evolutionary game theoretical studies: The outcome of cooperation depends on the evolutionary dynamics adopted , dictating how individual strategy evolves from generation to generation . Furthermore , evolutionary game dynamics on populations structured along multiple networks has been explored [33] , [34] , as well as the mechanisms which favor cooperation under adaptive population structures have been identified , both for non-repeated [35]–[43] and repeated games [44] , [45] . These results consubstantiate and keep stimulating an enormous amount of research work . Common to all these studies are the settings underlying the social dilemma: in the conventional view , every C pays a fixed cost c per game , providing the same benefit b to the partner . However , if what matters is the act of giving and not the amount given , then there is no reason to assume that everybody contributes the same cost c to each game . Depending on the amount of each individual contribution , the overall result of the evolutionary dynamics may change . The two person game introduced above provides not only the ideal ground to introduce such a diversity of contributions , but also an intuitive coupling between game dynamics and social embedding: The first ( second ) individual contributes a cost c1 ( c2 ) if playing C and nothing otherwise . Hence , player i ( i = 1 , 2 ) now gets the following payoff from this game: ( 2 ) reflecting the symmetry breaking induced by possibly different contributions from different cooperating individuals . This poses a natural question: Who will acquire an evolutionary edge under these conditions ? Often the amount that each individual contributes is correlated with the social context she is actually embedded in [28] , [46] , [47] . Modern communities are grounded in complex social networks of investment and cooperation , in which some individuals play radically different roles and interact more and more often than others . Empirical studies have demonstrated that social networks share both small-world properties and heterogeneous distribution of connectivities [48]–[50] . In such heterogeneous communities , where different individuals may be embedded in very different social environments , it is indeed hard to imagine that every C will always provide the same amount in every game interaction , hence reducing the problem to the standard two-person prisoner's dilemma studied so far . In the context of N-person games played in prototypical social networks , it has been found that the diversity of contributions greatly favors cooperation [28] . However , and similar to the relation between two-body and many-body interactions in the Physical Sciences , N-person public goods games have an intrinsic complexity which cannot be anticipated from two-person games: In the words of late William Hamilton , “The theory of many person games may seem to stand to that of two-person games in the relation of sea-sickness to a headache” [51] . Here , and besides the conventional scenario in which every C contributes the same cost c to each game she participates , we shall also explore the limit in which every C contributes the same overall amount c . However , this amount is shared between all games she participates , which are defined by the social network in which the players are embedded . For instance , c may be interpreted as the availability or the amount of resources each individual has to dedicate to all her commitments . Hence , the contribution to each game will depend now on the social context ( number of partners ) of each C , and heterogeneity will foster a symmetry breaking of pair-wise interactions , as two individuals may contribute different amounts to the same game . In this sense , cooperation will be identified with the act of giving and no longer with the amount given .
Figure 1 shows the final fraction of Cs for different classes of population structures and different contribution paradigms . At each time-step , every individual engages in a 2-person PGG with each of her neighbors . The accumulated payoff resultant from all interactions is associated with reproductive fitness or social success , which determines the behavior in the next generation [15] . We adopt the so-called pairwise comparison rule [52]–[54] for the social learning dynamics: Each individual copies the behavior of a randomly chosen neighbor with a probability which increases with the fitness difference ( see Methods for details ) . Figure 1a shows the outcome of evolving the conventional 2-person PD ( 1<F<2 ) , in which case each player contributes a fixed amount c to each game she participates . Different population structures are considered , one associated with a ( homogeneous ) regular network ( REG ) , the other with a ( strongly heterogeneous ) scale-free network ( SF ) . Real social networks fall somewhere between these limits [55] , and hence we also investigate a third class of population structure , represented by an exponential network ( EXP ) , exhibiting a level of heterogeneity intermediate between the previous two . The existence of a minority of highly connected individuals in SF networks ( line and circles ) allows the population to preserve high cooperative standards , while on homogeneous networks ( line and filled squares ) Ds dominate for the entire range of parameters [21] , [22] , as a result of the pairwise comparison rule adopted [56] . Heterogeneous networks thus pave the way for the emergence of cooperation . Highly connected individuals ( i . e . hubs ) work as catalysers of cooperative behavior , as their large number of interactions allows them to accumulate a high fitness . This , in turn , leads them to act as role models for a large number of social ties . To the extent that hubs are Cs , they influence the vast majority of the population to follow their behavior [23] . Clearly , this feature has a stronger impact on SF networks than on EXP networks , the difference between these two types of networks stemming from the presence or absence , respectively , of the preferential attachment mechanism . The results in Figure 1a are based on the assumption that each C contributes the same cost c to each game she plays – which we denote by conventional prisoner's dilemma ( CPD ) . This assumption is relaxed in Figure 1b where Cs now equally distribute the same cost c among all games they play – the regime we denote by distributed prisoner's dilemma ( DPD ) . Figure 1b shows what happens in this limit . While on homogeneous networks the fate of cooperation is the same as before − it amounts to rescaling of the intensity of selection − heterogeneity in the amount contributed by each individual to each game creates a remarkable boost in the final number of Cs for the entire range of F , which increases with increasing heterogeneity of the underlying network . Comparison with the results of Figure 1a shows that under DPD preferential attachment plays a prominent role , since it constitutes the network wiring mechanism distinguishing EXP networks from SF networks . Changing from CPD to DPD induces moderate boosts in the equilibrium fraction of Cs on EXP networks , but a spectacular boost of cooperation on SF networks: Hubs become extremely influential under DPD . In order to understand the mechanism underlying the population-wide boost of cooperation obtained , we consider a prototypical element of a heterogeneous network ( similarly to what has been done in [28] , [30] , [31] ) as shown in Figure 2 , and investigate the microscopic balance determining individual change . In particular , we investigate under which conditions the central C on the left – a stereotypical hub –becomes advantageous , that is , accumulates a higher fitness than any of her neighbors ( see Figure 2 ) . We consider a C-hub with z1 links ( k1 of which are Cs , left in Figure 2 ) and a D-hub with z2 links ( k2 of which are Cs , right in Figure 2 ) . We assume , for simplicity , that all neighbors of the C hub have z1L links each ( k1L of which are Cs ) , whereas all neighbors of the D hub have z2L links ( k2L of which are Cs ) . The remaining nodes have z0 links , where z0 stands , e . g . , for the average connectivity of the population . We implicitly assume that the neighbors of the hubs have smaller connectivities , and consequently we call them leaves . The conditions are explicitly provided in Figure 2 for both DPD and CPD . In both paradigms , for the C-hub to invade the D-hub ( or any of her D-leaf neighbors ) depends crucially on the difference between the number k1 of C-neighbors of the C-hub and the number k2 ( k1L ) of C-neighbors of the D-hub ( D-leaf ) . In both DPD and CPD the invasion threshold is always smaller for leaf invasion compared to hub invasion . Furthermore , the threshold for invasion is also smaller under DPD compared to CPD . Finally , as one would expect , all thresholds coincide when networks are homogeneous , the threshold conditions making it harder for invasion to occur in these networks . As a result , on heterogeneous networks , the conditions which render a C-hub advantageous with respect to a D-hub are more stringent than those associated with invasion of a neighbor D-leaf , which leads to an invasion pattern in which leaves are invaded before hubs [23] . Furthermore , one should not overlook that successful Ds tend to place other Ds in their neighborhood [23] which acts as a negative feedback mechanism reducing their fitness in time . On the contrary , successful Cs see their fitness increase in time , as more Cs join their neighborhood , reinforcing their fitness . The impact of the DPD paradigm , however , is most dramatic if one takes into consideration that the condition for the C-hub to become advantageous becomes less stringent the larger her connectivity . On the contrary , under the CPD paradigm , the cost of cooperation plays a major role in the overall fitness of the C-hub , which means that the larger her connectivity , the harder it will be for the C-hub to become advantageous with respect to any of her D-neighbors . Finally , the threshold conditions in Figure 2 also show that under DPD the range of game interaction is enlarged , as second neighbors of a hub also play a role in defining the invasion thresholds , unlike what happens under CPC . The insights provided by the prototypical configuration in Figure 2 become more explicit if one computes the outcome of cooperation in SF networks for social networks with different average connectivities and both contributive schemes ( Figure 3 ) . As the average degree ( z ) becomes sizable cooperation will inevitably collapse [21] , [23] , but while cooperation can hardly resist for z >10 in the case of CPD , under DPD Cs survive for values of z roughly four times larger . This is of particular importance given that social networks often exhibit high average connectivity values ( ) [48]: Cooperation prevails under a DPD contributive system , even on non-sparse static network structures . For intermediate regimes of heterogeneity ( EXP networks ) , under DPD cooperation is also sustained up to higher values of z , but to a lesser extent: Once more , the impact of large hubs resulting from the preferential attachment mechanism underlying SF networks plays an impressive role under DPD . The previous analysis allowed us to understand in which way heterogeneous networks , by inducing a symmetry breaking into the game dynamics , may favor cooperation . Furthermore , Figures 1 and 3 show how this indeed happens when one starts from initial conditions in which Cs occupy the nodes of a network with 50% probability . This approach , which is recurrent in numerical studies of evolutionary game dynamics , contrasts with the more conventional mean field analysis on which evolutionary game theory is rooted . There , the fact that all Cs and Ds in an infinite population have the same fitness , leads to a simple replicator equation in which the rate of change of Cs is proportional to a Gradient of selection G ( x ) , the interior roots of which dictate possible coexistence or coordination equilibria [14] . Here we shall define the finite population analog of G ( x ) , valid for any population size and structure ( see Methods ) . In doing so we overlook the microscopic details of the competition and self-organization of Cs and Ds , but we gain an overview of the game dynamics in a mean-field perspective . G becomes positive whenever cooperation is favored by evolution and negative otherwise . Whenever G = 0 , selection becomes neutral and evolution proceeds by random drift . Naturally , G will depend implicitly on the population structure , on the fraction x of Cs and also on how these Cs are spread in the network . In Figure 4 we plot G ( x ) as a function of x , for different values of F and different game paradigms ( CPD and DPD ) . Each configuration , here characterized by x , was generated assuming that each C ( D ) has , at least , one C ( D ) in her neighborhood , replicating the conditions observed in all numerical simulations . This is an important point , as strategy assortation constitutes a characteristic feature of evolutionary game dynamics in structured populations . Figure 4 shows that , unlike what happens on homogeneous networks , where Ds are always advantageous ( not shown ) , SF networks effectively transform a prisoner's dilemma into a different game . Figure 4a indicates that , in the case of CPD , introducing diversity in roles and positions in the social network effectively leads to a coordination game [57] , [58] , characterized ( in an infinite , well-mixed population ) by a critical fraction x* above which Cs are always advantageous ( G<0 for x<x* and G>0 for x>x* ) . This result provides a powerful qualitative rationale for many results obtained previously on heterogeneous networks under strong selection [21] , [22] , [28] in which degree heterogeneity is shown to induce cooperative behavior , inasmuch as the initial fraction of Cs is sufficient to overcome the coordination threshold . Moreover , Figure 4b shows that changing the contributive scheme from CPD to DPD in SF population structures acts to change a prisoner's dilemma effectively into a Harmony game where Cs become advantageous irrespectively of the fraction of Cs ( x*≈0 ) .
The present study puts in evidence the impact of breaking the symmetry of cooperative contributions to the same game . On strongly heterogeneous networks , the results of Figures 1b and 3 provide an impressive account of the impact of this diversity of contributions . Overall , our results strongly suggest that whenever the act of cooperation is associated to the act of contributing , and not to the amount contributed , cooperation blooms inasmuch as the structure of the social web is heterogeneous , leading individuals to play diverse roles . The multiplicity of roles and contributions induced by the social structure effectively transforms a local cooperative dilemma into a global coordination game [57] . The latter embodies an exemplary representation of the social contracts [57] found in several instances of animal [59] , [60] and human [61] , [62] collective dilemmas . This work provides additional evidence that , while locally , cooperation can be understood as a prisoner's dilemma , globally , the possibilities opened by the intricate nature of collective dynamics of cooperation [63] often lead to a dynamical portrait that is effectively described by a coordination dilemma instead of a defection dominance dilemma [57] .
Each individual is assigned to a node of a network , whereas interactions are represented by links between nodes . In each generation , all pairs of individuals directly connected , engage in a single round of the game . As usual , the accumulated payoff from all interactions emulates the individual fitness ( fi ) or social success and the most successful individuals will tend to be imitated by their neighbors . Such behavioral evolution is implemented using the pairwise comparison rule [52] , [54]: at each time step an individual x will adopt the strategy of a randomly chosen neighbor y with a probability given by the ubiquitous Fermi distribution from statistical physics [52] , [54] , in which β , the inverse temperature in Physics , translates here into noise associated with errors in decision making . For high values of β we obtain the imitation dynamics commonly used in cultural evolution studies whereas for β≪1 evolution proceeds by random drift . The strong selection regime that we adopt here ( β = 10 . 0 ) enhances both the influence of the payoff values in the individual fitness and the role played by the social network . It is noteworthy that a detailed study of the impact of β on game dynamics on heterogeneous networks is still lacking , unlike what happens on homogeneous networks [52] , [54] , [64] . The results in Figures 1 and 3 were obtained for populations of N = 103 individuals starting with 50% of Cs randomly distributed on the network . In all cases we used the value c = 1 for the cost of cooperation . The scale-free networks were generated using a direct implementation of the Barabási-Albert ( BA ) model , based on growth and preferential attachment [65] , whereas exponential networks were generated replacing the preferential attachment by uniform attachment in the previous model [49] . Different mechanisms could be used [38] , [42] , [48] , [66]–[68] to generate SF degree distributions portraying features not present in the BA model . In general , however , SF networks lead to evolutionary dynamical behaviors which are similar to those observed in BA networks [24] , [27] , [42] , [68]–[70] , which may also depend on the way individual fitness is defined [23] , [29] , [71] , [72] . The equilibrium fraction of Cs results from averaging over 2000 generations after a transient period of 105 generations and each point in Figures 1 and 3 corresponds to an average over 103 runs and networks . The results are independent from the updating strategy ( synchronous , asynchronous ) , population size ( N >500 ) and robust to the existence of a small number of mutations in each time-step . In Figure 4 , gradients of selection were obtained by calculating , where is the average frequency of transitions increasing ( decreasing ) the number of Cs for each random configuration with xN Cs . G ( x ) represents a finite population analogue ( using the pairwise comparison rule [52] , [54] ) of the gradient of selection in infinite well-mixed populations [14] , where and are the fitness values of Cs and Ds . Each value was obtained by averaging over 105 different randomly generated configurations and networks .
|
Humans contribute to a broad range of cooperative endeavors . In many of them , the amount or effort contributed often depends on the social context of each individual . Recent evidence has shown how modern societies are grounded in complex and heterogeneous networks of exchange and cooperation , in which some individuals play radically different roles and/or interact more than others . We show that such social heterogeneity drastically affects the behavioral dynamics and promotes cooperative behavior , whenever the social dilemma perceived by each individual is contingent on her/his social context . The multiplicity of roles and contributions induced by realistic population structures is shown to transform an initial defection dominance dilemma into a coordination challenge or even a cooperator dominance game . While locally defection may seem inescapable , globally there is an emergent new dilemma in which cooperation often prevails , illustrating how collective cooperative action may emerge from myopic individual selfishness .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"evolutionary",
"biology/human",
"evolution",
"mathematics"
] |
2009
|
Population Structure Induces a Symmetry Breaking Favoring the Emergence of Cooperation
|
Somatic mutations of the cohesin complex subunit STAG2 are present in diverse tumor types . We and others have shown that STAG2 inactivation can lead to loss of sister chromatid cohesion and alterations in chromosome copy number in experimental systems . However , studies of naturally occurring human tumors have demonstrated little , if any , correlation between STAG2 mutational status and aneuploidy , and have further shown that STAG2-deficient tumors are often euploid . In an effort to provide insight into these discrepancies , here we analyze the effect of tumor-derived STAG2 mutations on the protein composition of cohesin and the expected mitotic phenotypes of STAG2 mutation . We find that many mutant STAG2 proteins retain their ability to interact with cohesin; however , the presence of mutant STAG2 resulted in a reduction in the ability of regulatory subunits WAPL , PDS5A , and PDS5B to interact with the core cohesin ring . Using AAV-mediated gene targeting , we then introduced nine tumor-derived mutations into the endogenous allele of STAG2 in cultured human cells . While all nonsense mutations led to defects in sister chromatid cohesion and a subset induced anaphase defects , missense mutations behaved like wild-type in these assays . Furthermore , only one of nine tumor-derived mutations tested induced overt alterations in chromosome counts . These data indicate that not all tumor-derived STAG2 mutations confer defects in cohesion , chromosome segregation , and ploidy , suggesting that there are likely to be other functional effects of STAG2 inactivation in human cancer cells that are relevant to cancer pathogenesis .
Cohesin is a multiprotein complex comprised of four primary subunits ( SMC1A , SMC3 , RAD21 , and either STAG1 or STAG2 ) and four regulatory subunits ( WAPL , CDCA5 , and PDS5A or PDS5B ) that is responsible for sister chromatid cohesion , regulation of gene expression , DNA repair , and other phenotypes [1 , 2] . Somatic mutations of cohesin subunits are common in a wide range of pediatric and adult cancers [3 , 4] . STAG2 ( also known as SA2 ) is the most commonly mutated subunit , presumably in part because the STAG2 gene is located on the X chromosome and therefore requires only a single mutational event to be inactivated [5] . Approximately 85% of tumor-derived STAG2 mutations lead to premature truncation of the encoded protein , whereas approximately ~15% are missense mutations . STAG2 mutations are particularly common in bladder cancer ( present in 30–40% of the most common non-muscle invasive tumors ) , Ewing sarcoma ( present in ~25% of tumors ) , and myeloid leukemia ( present in ~8% of tumors ) , and are also present in glioblastoma multiforme ( GBM ) , melanoma , and other tumor types [6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15] . Highlighting the importance of STAG2 as a cancer gene , in 2014 The Cancer Genome Atlas identified STAG2 as one of only 12 genes that are significantly mutated in four or more human cancer types ( the others were TP53 , PIK3CA , PTEN , RB1 , KRAS , NRAS , BRAF , CDKN2A , FBXW7 , ARID1A and KMT2D; [16] ) . Among the other components of cohesin , RAD21 is the most commonly mutated subunit , with mutations of SMC1A and SMC3 also present in a subset of tumors . In addition to the frequent mutations in human tumors , the role of STAG2 inactivation in cancer pathogenesis is also highlighted by the fact that it is commonly altered in transposon-mediated tumorigenesis in mouse model systems [17 , 18] . The mechanism ( s ) through which cohesin gene mutations confer a selective advantage to cancer cells is controversial . In our initial studies identifying STAG2 mutations in cancer , we demonstrated using isogenic human cultured cell systems that STAG2 mutations can lead to alterations of chromosome counts and aneuploidy [5 , 6] . These findings were consistent with other observations in yeast , mice , and other model systems indicating that mutations in cohesin subunits lead to chromosomal non-disjunction and aneuploidy [19 , 20 , 21 , 22] . However , more recent studies of naturally occurring human tumors have demonstrated either weak or no correlations between the presence of cohesin gene mutations and aneuploidy [8] . Furthermore , a subset of naturally occurring human tumors harboring cohesin gene mutations are euploid . These conflicting data are likely in part attributable to the paucity of currently available functional data on tumor-derived STAG2 mutations . For example , our reported functional studies of tumor-derived STAG2 mutations have been primarily limited to two truncating mutations present in the human H4 and 42MGBA GBM cell lines ( which were corrected by human somatic cell gene targeting ) . Others have demonstrated that myeloid leukemia cell lines harboring specific cohesin gene mutations have reduced levels of chromatin-bound cohesin compared to non-isogenic myeloid leukemia cells harboring wild type cohesin genes [10] . These authors further showed that ectopic expression of wild-type STAG2 can lead to growth suppression of myeloid leukemia cells harboring specific endogenous mutations of STAG2 . Other work has focused on the functional consequences of STAG2 depletion in mammalian cells , without testing individual mutations . For example , it has recently been shown that transient depletion of STAG2 in human cells leads to a significant increase in errors in chromosome segregation due to aberrant kinetochore-microtubule attachments [23] . Furthermore , we and others have shown that stable depletion of STAG2 in bladder cancer cells leads to alterations in chromosome counts [6 , 24] . Here we report a comprehensive analysis of the effects of tumor-derived STAG2 mutations on the composition of cohesin and on mitotic phenotypes generally attributed to cohesin inactivation . We find that many mutant STAG2 proteins retain the ability to interact with cohesin , but result in a generalized reduction in levels of core cohesin subunits and in the ability of PDS5A , PDS5B , and WAPL to interact with the core cohesin ring . We then created and studied isogenic sets of human cells in which the endogenous allele of STAG2 was modified via the addition of nine different tumor-derived mutations . While all nonsense mutations led to abrogation of sister chromatid cohesion and some nonsense mutations led to anaphase defects , tumor-derived missense mutations behaved like wild-type in these assays . Furthermore , few of the cell lines demonstrated the expected alterations in chromosome counts .
To express tumor-derived STAG2 mutations in human cells , we created a full length human STAG2 expression vector with a 1x FLAG/Streptavidin Binding Peptide ( SBP ) dual epitope tag at the amino terminus . This human STAG2 expression vector corresponds to transcript CCDS43990 . This new cDNA expression vector differs from the STAG2 expression vector used in previous studies ( transcript CCDS14607 ) , which was a naturally-occurring splice variant missing the 37 amino acids encoded by exon 30 . Transfection of the new , dual epitope-tagged expression vector into 293T cells led to expression of the tagged STAG2 cDNA ( Fig 1 ) . In naturally occurring human tumors , ~85% of STAG2 mutations are truncating and ~15% are missense , with mutations spread roughly evenly throughout the gene [3] . There is also a minor mutational hotspot in Ewing sarcoma ( R216X; ref . 13 ) , found in ~20% of Ewing sarcoma tumors harboring mutations of STAG2 . To model these mutations , we used site-directed mutagenesis to introduce 50 tumor-derived mutations into the cloned cDNA . The mutations were derived from numerous tumor types and were spread evenly throughout the gene ( Fig 1A and S1 Table ) . These constructs were transfected into 293T cells , and their expression measured by Western blotting with FLAG antibodies . As depicted in Fig 1B , most of the truncating mutations were expressed very poorly , possibly due to nonsense mediated decay of the encoded transcript . Four missense mutations ( #4 , 5 , 22 , 44 ) were expressed at levels equivalent to wild-type protein . Arguably the most straightforward hypothesis regarding the loss of activity caused by tumor-derived STAG2 mutations is that the mutations abolish the ability of the encoded STAG2 protein to interact with the rest of the cohesin complex . To test this , we measured the interaction of three missense mutations and two nonsense mutations with cohesin subunits SMC1A , SMC3 , RAD21 , STAG2 , STAG1 , WAPL , and PDS5A by immunoprecipitation ( IP ) —Western blotting . As depicted in Fig 2 , each of the three proteins encoded by STAG2 genes harboring tumor-derived missense mutations retained its ability to interact with cohesin . Similarly , a truncated protein encoded by a STAG2 gene harboring a tumor-derived late truncating mutation ( S1075X ) also retained its ability to interact with cohesin . In contrast , the STAG2 protein encoded by a gene with an earlier truncating mutation ( S653X ) lost the ability to interact with the rest of the cohesin complex . We next determined the boundary in STAG2 protein at which tumor-derived truncating mutations retained their ability to interact with cohesin . To do this , we created nine additional tumor-derived truncating mutations in STAG2 from amino acid 1021 to amino acid 1137 in the cloned STAG2 expression vector ( #51–59; Fig 3A , S1 Table ) , transfected 293T cells , and performed IP-Western blotting . As depicted in Fig 3B , amino acids 983 to 1268 are dispensable for the interaction with cohesin , corresponding roughly to the beginning of the regulatory phosphorylation sites in the STAG2 protein [25] . These results are concordant with data from the recently published crystal structure of STAG2 , which shows that STAG2 interacts with cohesin via an extensive interface with RAD21 that spans nearly the entire length of STAG2 and that is not disrupted by missense mutations targeting the interface [26] . Taken together , these data demonstrate that tumor-derived mutations of STAG2 do not uniformly lose the ability to interact with cohesin , indicating that at least some tumor-derived mutations of STAG2 must affect a key function of STAG2 other than its ability to interact with cohesin . Having shown that many STAG2 mutations do not abrogate the ability of STAG2 to interact with cohesin ( Figs 2 and 3 ) , we next hypothesized that tumor-derived mutations in STAG2 might instead lead to generalized abnormalities in the subunit composition of the cohesin complex itself . To test this , we studied previously described isogenic sets of H4 and 42MGBA GBM cells in which the endogenous naturally occurring mutant allele of STAG2 were corrected by somatic cell gene targeting [5] . Non-detergent nuclear lysates were prepared from these isogenic sets of cells , endogenous cohesin was immunoprecipitated using SMC3 antibodies , and Western blot analysis performed using antibodies specific to the cohesin components SMC1A , SMC3 , RAD21 , STAG1 , STAG2 , WAPL , PDS5A , and PDS5B . Correction of the endogenous mutant allele of STAG2 led to an increase in total levels of the core cohesin subunits SMC1A , SMC3 , and RAD21 in both H4 and 42MGBA cells , but did not affect the total levels of regulatory factors WAPL , PDS5A , or PDS5B ( Fig 4A ) . Targeted correction of mutant STAG2 did not affect the ability of SMC3 to co-immunoprecipitate SMC1A , SMC3 , or RAD21 ( Fig 4B; S1 Fig ) , demonstrating that STAG2 is not required for assembly of the core cohesin ring . However , targeted correction of mutant STAG2 substantially enhanced the ability of WAPL , and to a lesser extent PDS5A and PDS5B , to interact with cohesin ( Fig 4B ) , consistent with recently published structural studies showing that STAG2 functions in part as a structural scaffold for the interaction of WAPL with the core cohesin ring [27] . Taken together , these data indicate that the presence of tumor-derived STAG2 mutations results in a decrease in levels of cohesin and a reduction in ability of WAPL , PDS5A , and PDS5B to interact with cohesin . We next wanted to determine the functional significance of a representative subset of these mutations in human cells . In an initial attempt to do this , we created several STAG2 expressing lentiviruses for reconstituting wild-type and mutant STAG2 expression in cancer cells harboring STAG2 mutations . However , we found that expression of STAG2 from these ectopic systems was inconsistent and substantially less than that expressed from the endogenous allele in cells with wild-type STAG2 . To circumvent this , AAV-mediated human somatic cell gene targeting was used to “knock-in” ( KI ) tumor-derived mutations into the endogenous allele of STAG2 in HCT116 cells , a human cancer cell line with a single , wild-type allele of STAG2 , intact sister chromatid cohesion , and a near-diploid karyotype . Of note , we have previously used a similar approach to KI a single , non-tumor derived amino terminal nonsense mutation at codon six into STAG2 in HCT116 cells and demonstrated a substantial reduction in sister chromatid cohesion and alterations in chromosome counts [5] . To do this , we created nine AAV-based targeting vectors in the pAAV-SEPT vector system and used these vectors to create clonal derivatives of HCT116 cells in which the single endogenous allele of STAG2 had been mutated ( Fig 5A–5C ) . The details of this approach are described in Materials and Methods . Targeting efficiencies were 14–29% ( S2 Table ) . Western blot analysis with a monoclonal antibody recognizing the carboxyl terminus of STAG2 demonstrated that , as expected , cells harboring the seven nonsense mutations completely lacked expression of the c-terminal epitope ( Fig 5D ) . In contrast , the two missense mutations were expressed at levels equivalent to wild-type STAG2 . Western blotting with an antibody recognizing the amino terminus of STAG2 further demonstrated that the truncated proteins were expressed poorly , concordant with the ectopic expression data shown in Fig 1 and consistent with the notion that the transcripts may have been degraded via nonsense-mediated decay . To determine whether the introduction of tumor-derived mutations had an effect on the proliferation of HCT116 cells , in vitro cellular proliferation assays were performed on the isogenic sets of cells . As shown in Fig 6A , mutation of STAG2 had a slightly adverse effect on the proliferation of HCT116 cells . To confirm and extend this result , we performed similar assays on two isogenic sets of GBM cells ( 42MGBA , H4 ) in which the endogenous mutant allele of STAG2 had been corrected by gene targeting . As predicted by the result in the HCT116 cell lines , in both H4 cells and 42MGBA cells , correction of STAG2 led to slightly enhanced proliferation ( Fig 6B and 6C ) . Taken together , these data suggest that unlike in the case of most tumor suppressor genes , mutational inactivation of STAG2 may have an adverse effect on proliferation . To measure the effects of tumor-derived nonsense and missense mutations on sister chromatid cohesion , we enriched the HCT116 cells and STAG2 KI derivatives in mitosis by short treatment with nocodazole , then examined prometaphase chromosome spreads to analyze sister chromatid cohesion . For details , see Materials and Methods . As shown in Fig 7 , all tumor-derived nonsense mutations in STAG2 led to a reduction in the integrity of sister chromatid cohesion . In contrast , tumor-derived missense mutations displayed wild-type sister chromatid cohesion . These data suggest either that STAG2 nonsense mutations have a different pathogenic mechanism than STAG2 missense mutations , or that abrogation of sister chromatid cohesion is not the key property targeted by tumor-derived mutations in STAG2 . Next , we tested whether cells harboring tumor-derived mutations in STAG2 showed a reduction in the integrity of sister chromatid segregation during anaphase . To do this , proliferating cells were fixed , triple stained with DAPI and antibodies to α-tubulin and the centromere ( ACA ) , and the fractions of anaphase cells with chromosome missegregation errors was determined ( for details , see Materials and Methods ) . Three of seven cell lines harboring STAG2 truncating mutations ( S653X , S1075X , and S1215X ) had a statistically significant increase in lagging chromosomes compared to isogenic cells with wild-type STAG2 ( Fig 8; S3 Table ) . There was no statistically significant increase in the fractions of cells with DNA bridges , acentric fragments , or multipolar anaphases in any of the STAG2 mutant cells ( S3 Table ) . In two previous studies , we found that introduction of a non-tumor derived mutation or stable lentiviral depletion of wild-type STAG2 led to alterations in chromosome counts in human cells–in both cases increasing the modal chromosome number by one [5 , 6] . To determine if similar results would be obtained after introduction of tumor-derived mutations into STAG2 , chromosomes were counted in parental HCT116 cells and derivatives with each of the nine tumor-derived mutations . As described in Materials and Methods , prometaphase chromosome spreads were prepared , and the number of chromosomes in 100 cells was determined for each cell line . Surprisingly , there was no change in modal chromosome number after introduction of eight of the tumor-derived mutations , with the exception of the introduction of S653X , which led to an increase in the modal chromosome number by one ( Table 1 ) . This experiment indicates that many tumor-derived mutations in STAG2 do not have an overt effect on chromosome count when introduced into HCT116 cells .
Somatic mutations of genes encoding components of the cohesin complex are common in a variety of adult and pediatric human cancers . The most commonly mutated cohesin gene is STAG2 , with other cohesin genes such as RAD21 , SMC1A and SMC3 also mutated in a subset of tumors . Now that the frequencies and tumor-type specificities of these mutations have been established , it has become increasingly important to determine the functional role played by these mutations in tumorigenesis . In the studies presented here we have focused our attention on the role of tumor-derived mutations in STAG2 on the protein composition of cohesin , cellular proliferation , mitosis , and aneuploidy . Arguably the most straightforward hypotheses regarding the loss of function caused by tumor-derived STAG2 mutations is that the mutations either abolish the ability of STAG2 to interact with the rest of the cohesin complex , or that they lead to more generalized abnormalities in the subunit composition of cohesin . To test these hypotheses , we expressed 59 tumor-derived mutations in human cells in an epitope tagged STAG2 cDNA . Many of the mutations led to virtually complete loss of expression of the encoded protein , consistent with the expectation that STAG2 genes harboring tumor-derived nonsense mutations produce transcripts that are degraded by nonsense mediated decay . We then performed IP-Western blot analyses on the mutant STAG2 proteins that were expressed , and showed that many tumor-derived mutations do not abolish the ability of the encoded protein to interact with cohesin . These data indicate that the ability to interact with the rest of the cohesin complex is not—or at least not the only—key property of STAG2 abrogated by tumor-derived mutations in the gene . We then tested an alternate hypothesis—that STAG2 mutations cause alterations in the amount and/or subunit composition of cohesin in human cancer cells . We found that cells harboring STAG2 mutations have roughly half as much SMC1A , SMC3 , and RAD21 as otherwise isogenic cells with wild-type STAG2 . This result is consistent with previous studies showing that myeloid leukemias with STAG2 mutations have less chromatin-bound cohesin than leukemias with wild-type STAG2 [10] . Our result suggests that this may have been due to a reduction in cohesin itself rather than a reduction in the efficiency by which cohesin binds chromatin . Next , we tested the hypothesis that mutations in STAG2 result in alterations in the subunit composition of cohesin . We found that STAG2 mutations did not affect the composition of the core cohesin ring ( composed of SMC1A , SMC3 , RAD21 , and STAG1 or STAG2 ) . However , the presence of a STAG2 mutation did adversely affect the ability of cohesin regulatory factor WAPL ( and to a lesser extent PDS5A and PDS5B ) to interact with the core cohesin ring . These results are consistent with current models of cohesin structure , which suggest that STAG2 is not itself a key structural scaffold for assembly of the core cohesin ring , but instead is tethered to cohesin via its interaction with RAD21 . Furthermore , our finding that interaction of WAPL with cohesin depends in part on STAG2 is in agreement with recently published structural studies showing that STAG2 functions as a structural scaffold for the interaction of WAPL with the core cohesin ring [27] . Of note , WAPL is required for removal of cohesin from chromosome arms during prophase , and reductions in cohesin-bound WAPL have previously been shown to lead chromosomal non-disjunction and aneuploidy [28] . Next , we used human somatic cell gene targeting to introduce a subset of these tumor-derived mutations into the endogenous allele of STAG2 in a chromosomally stable , near diploid human cell line . Despite the fact that only 14% of tumor-derived mutations in STAG2 are missense ( the remainder are truncating ) , we intentionally introduced several missense mutations since we believe their effects are likely to be more subtle and therefore could shed important light on the specific functional deficiencies of STAG2 in cancer . Interestingly , genetically modified cells harboring STAG2 mutations proliferated slightly more slowly than isogenic parental cells with wild-type STAG2 –a result which was then confirmed by measuring the proliferation of several additional isogenic sets of GBM cells in which the endogenous mutant allele of STAG2 had been corrected by human somatic cell gene targeting . This observation that endogenous STAG2 mutations slow cellular proliferation is surprising , since mutations in tumor suppressor genes are generally thought to enhance cellular proliferation . Of note , our proliferation data are in disagreement with recent results from Balbas-Martinez et al . and Kon et al . , who showed that ectopic re-expression of wild-type STAG2 and RAD21 in cancer cells harboring endogenous mutations of those genes leads to suppression of proliferation [8 , 10] . It is possible that the discrepancies between our data and those of Balbas-Martinez and Kon et al . are due to differences between ectopic and endogenous expression of wild-type STAG2 and/or cell type specific differences . Using these isogenic sets of cells , we next tested the effect of tumor-derived nonsense and missense mutations in STAG2 on sister chromatid cohesion , mitotic fidelity , and chromosome counts . As expected based on our previous studies , all seven nonsense mutations in STAG2 led to substantial reductions in the integrity of sister chromatid cohesion during metaphase . However , to our surprise , tumor-derived missense mutations retained wild-type levels of sister chromatid cohesion . Furthermore , only a subset of the nonsense mutations ( and none of the missense mutations ) caused an increase in lagging chromosomes during anaphase , generally considered a marker of chromosomal instability . Furthermore , despite the substantial defects in cohesion in cells with STAG2 nonsense mutations ( and the increase in lagging chromosomes in a subset of the cell lines ) , only one of the mutant STAG2 KI cell lines demonstrated alterations in chromosome counts when compared to isogenic cells with wild-type STAG2 genes . These results raise a number of interesting issues and questions regarding the role of STAG2 gene mutations in cancer pathogenesis . First , we believe that the study of tumor-derived missense mutations provides a particularly useful window into the specific functions of STAG2 that are relevant to cancer pathogenesis , since the adverse effects of missense mutations are likely to more specific to cancer pathogenesis than the more pleiotropic effects of early truncating mutations ( with the caveat that it is a formal possibility that a subset of missense mutations are passenger mutations ) . The fact that the missense mutations tested retained wild-type cohesion , mitotic integrity , and ploidy suggests that perhaps these phenotypes are not central to the cancer-causing effects of STAG2 mutations . Such an interpretation could help resolve a fundamental current discrepancy in the field–that all tumor-derived mutations tested so far ( all of which have been truncating mutations ) result in dramatic reductions in cohesion , and yet many naturally occurring tumors with STAG2 mutations appear to be euploid . Alternatively , it is a formal possibility that nonsense and missense mutations in STAG2 lead to deficiencies in two different functions of STAG2 and therefore lead to cellular transformation through two different pathways . However , we consider this possibility to be unlikely . The chromosome counting data presented here are especially surprising in light of our previously published data showing that the introduction of an amino terminal , non-tumor derived truncating mutation into STAG2 leads to an increase in modal chromosome number [5] . Similar results were also obtained in a bladder cancer cell line with shRNA-mediated depletion of STAG2 [6] . It is particularly surprising that only one of the seven nonsense mutations tested resulted in an alteration in chromosome count , despite the fact that all nonsense mutations led to substantial reductions in sister chromatid cohesion and a subset led to an increase in lagging chromosomes during anaphase . Though surprising in light of the preponderance of published data on the relationship of cohesion to aneuploidy , these data are consistent with published work in Saccharomyces cerevisiae suggesting that low levels of cohesin can be sufficient to maintain euploidy [29 , 30] . When taken together with the results presented for the missense mutations tested , these results further call into question whether the cancer relevant phenotypes of STAG2 mutations are directly related to cohesion and aneuploidy . In summary , here we analyze the properties of a substantial number of tumor-derived mutations in STAG2 . In addition to evaluating the effect of the mutations on the composition of cohesin , we demonstrate that the mutations do not uniformly abrogate sister chromatid cohesion , anaphase integrity , or the ability of the cells to control their ploidy . These data suggest either that different mutations of STAG2 have different mechanisms through which they cause cancer , or that the specific mechanism ( s ) through which STAG2 and other cohesin gene mutations contribute to cancer pathogenesis remain , at present , unknown .
A human STAG2 cDNA ( CCDS43990 ) with an amino terminus 1X FLAG/SBP dual epitope tag was synthesized de novo by Genscript and cloned into pUC57 . Tumor-derived mutations were created by site-directed mutagenesis using the QuikChange II XL kit ( Stratagene ) as directed by the manufacturer . Wild-type and mutant cDNAs were then subcloned into pcDNA3 . 1 . HCT116 cells were obtained from ATCC . H4 and 42MGBA parental cells and gene targeted derivatives in which the endogenous mutant allele of STAG2 was corrected by AAV-mediated human somatic cell gene targeting were described previously [5] . Primary antibodies for immunoblotting were FLAG ( M2 ) from Sigma Aldrich; STAG2 ( J-12; carboxyl-terminus epitope ) , STAG1 ( A-9 ) , RAD21 ( B-2 ) , SMC3 ( E-3 ) , SMC1A ( M-16 ) , WAPL ( A-7 ) , and PDS5A ( S-20 ) from Santa Cruz Biotechnology; and PDS5B ( A300-537A ) and STAG2 ( A302-580A; amino terminus epitope ) from Bethyl Laboratories . Antibodies and conjugated beads for immunoprecipitation were FLAG-M2 beads from Sigma Aldrich , SMC3 ( A300-060A ) from Bethyl Laboratories , and Streptavidin Plus UltraLink Resin from Pierce . Antibodies for immunofluorescence were anti-centromere antibody ( ACA ) from Geisel School of Medicine , α-tubulin ( DM1α ) from Sigma Aldrich , and Alexa Fluor 488 and 568 from Molecular Probes . Protein lysates for direct Western blotting were prepared in NP40 lysis buffer ( 50 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 1% NP40 ) . Nuclear lysates used for affinity purification were prepared using a modification of Dignam's nondetergent lysis method [31] . Protein concentrations were determined using the bicinchoninic assay ( Pierce ) . For affinity purification of proteins tagged with Streptavidin Binding Peptide ( SBP ) , streptavidin beads ( Pierce ) were washed once with Tris-buffered saline ( TBS; 150 mM NaCl ) and then incubated with protein lysates derived from cells transfected with constructs expressing epitope-tagged proteins . Samples were incubated with rotation at 4°C for 1 h . Beads were then washed three times with TBS , and bound proteins eluted by boiling in sample buffer , separated by SDS-PAGE , and Western blot analysis performed . Homology arms ( ~1 kb each ) for creation of STAG2 KI vectors were synthesized by Genscript and cloned into the pAAV-SEPT-Acceptor vector [32] . Next , transient stocks of AAV-2 virions were created by co-transfection of 293T cells with STAG2 KI vectors together with pAAV-RC ( Stratagene ) and pHELPER ( Stratagene ) using X-tremeGENE 9 ( Roche ) . Two days after transfection , media was aspirated and cell monolayers were scraped into 1 mL PBS and subjected to four cycles of freeze/thaw . The lysate was then clarified by centrifugation at 12 , 000 rpm for 10 min in a benchtop microfuge to remove cell debris , and the virus-containing supernatant was aliquoted and stored at −80°C . 200 μL of virus was then used to infect HCT116 cells in T25 tissue culture flask , and cells were passaged at limiting dilution into 96-well plates in the presence of 1 . 0 mg/mL G418 . Individual G418-resistant clones were expanded and used for the preparation of genomic DNA . Clones were tested for homologous integration of the targeting vector using a primer pair specific for the targeted allele , and integration of the targeted mutation was confirmed by DNA sequencing . For missense mutations , KI cells were then infected with a Cre-expressing adenovirus to remove the FLOXed splice acceptor-IRES-neoR gene . Cells were seeded at either 500 or 1000 cells/well in 96 well plates and cell number measured every 48 hours using the Cell Titre Glo Assay ( Promega ) . Cells were grown to approximately 80% confluence , then treated with 330 μM nocodazole for one hour . Cells were then harvested by trypsinization and hypotonically swollen in 40% medium/60% Vienna tap water for 5 min at room temperature . Cells were fixed with freshly made Carnoy’s solution ( 75% methanol , 25% acetic acid ) , and the fixative was changed several times . For spreading , cells in Carnoy’s solution were dropped onto glass slides and dried at 37C . Slides were stained with 5% Giemsa at pH 6 . 8 for 7 min , washed briefly in tap water , air dried , and mounted with Entellan mounting medium . Each experiment was performed independently twice . Cells were fixed with 3 . 5% paraformaldehyde ( Alfa Aesar ) for 10 min , washed with Tris-buffered saline with 5% bovine serum albumin ( TBS-BSA ) and 0 . 5% Triton X-100 for 2 x 5 min with vigorous shaking and then rinsed in TBS-BSA . Primary antibodies were diluted in TBS-BSA containing 0 . 1% Triton X-100 and incubated for 1–3 h at room temperature . Cells were then washed with TBS-BSA/0 . 1% Triton X-100 for 3 x 10 min with vigorous shaking . Secondary antibodies were diluted in TBS-BSA plus 0 . 1% Triton X-100 and coverslips were incubated for 1–2 h at room temperature . Cells were then washed with TBS-BSA containing 0 . 1% Triton X-100 with DAPI for 3 x 10 min with vigorous shaking and then mounted using ProLong Gold antifade reagent ( Molecular Probes ) . Images were acquired with a cooled charge-coupled device camera ( Andor Technology , Belfast , UK ) mounted on a Nikon Eclipse Ti microscope ( Nikon , Melville , NY ) with a 60X , 1 . 4 numerical aperture objective . Image series in the z-axis were obtained using 0 . 2-μm optical sections . Image deconvolution and contrast enhancement was performed using AutoQuant X3 ( Media Cybernetics ) , Elements software ( Nikon ) , Image J and Adobe Photoshop software . Final images represent selected overlaid planes . For quantifications of anaphase segregation defects , chromatids were counted as lagging if they contained centromere staining ( using ACA ) or acentric if they did not contain centromere staining , in the spindle midzone separated from centromeres/kinetochores at the poles . DNA bridges were counted when a piece of stretched DNA ( visualized by DAPI staining ) spanned the area between the two newly formed daughter nuclei in anaphase and no centromere staining was evident . Tripolar anaphases were counted when three clear chromosome populations ( with centromere staining ) were evident in anaphase . For quantification of anaphase chromosome missegregation rates , no fewer than 150 anaphases per cell line were quantified . Statistical analysis for anaphase chromosome missegregation rates was performed using Fisher’s exact two-tailed test . Cultured cells were treated with 0 . 02 μg/ml colcemid for 45 minutes at 37°C . The cells were then trypsinized , centrifuged at 200 x g , and the cell pellet resuspended in warmed hypotonic solution and incubated at 37°C for 11 minutes . The swollen cells were then centrifuged and the pellet resuspended in 8 mL of Carnoy’s fixative ( 3:1 methanol:glacial acetic acid ) . After incubation in fixative at room temperature for 22 minutes , the cell suspension was centrifuged and washed twice in Carnoy’s fixative . After the last centrifugation , the cells were resuspended in 1 to 3 mL freshly prepared fixative to produce an opalescent cell suspension . Drops of the final cell suspension were placed on clean slides and air-dried . Slides were stained with a 1:3 mixture of Wright’s stain and 60 mM phosphate buffer for 4–10 minutes , washed with tap water for 5 seconds , and then air-dried . Chromosomes were counted in 100 prometaphases per cell line .
|
Mutations of the STAG2 gene are common in several types of adult and pediatric cancers . In fact , STAG2 is one of only 12 genes known to be significantly mutated in four of more types of cancer . The STAG2 gene encodes a protein component of the “cohesin complex , ” a ring-like structure that binds chromosomes together ( e . g . , “coheres” them ) until the cohesin complex is degraded during cell division , allowing replicated chromosomes to separate normally to the two new cells . The cohesin complex also plays important roles in other cellular processes including turning genes on and off , and in repairing damaged genes . Here we analyze the effect of cancer-causing mutations in STAG2 on its ability regulate the separation of chromosomes during cell division .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"anaphase",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"chromatids",
"mutation",
"immunoprecipitation",
"nonsense",
"mutation",
"research",
"and",
"analysis",
"methods",
"missense",
"mutation",
"chromosome",
"biology",
"chromosomal",
"disorders",
"precipitation",
"techniques",
"clinical",
"genetics",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"chromosomes"
] |
2016
|
Intact Cohesion, Anaphase, and Chromosome Segregation in Human Cells Harboring Tumor-Derived Mutations in STAG2
|
Cancer development is driven by series of events involving mutations , which may become fixed in a tumor via genetic drift and selection . This process usually includes a limited number of driver ( advantageous ) mutations and a greater number of passenger ( neutral or mildly deleterious ) mutations . We focus on a real-world leukemia model evolving on the background of a germline mutation . Severe congenital neutropenia ( SCN ) evolves to secondary myelodysplastic syndrome ( sMDS ) and/or secondary acute myeloid leukemia ( sAML ) in 30–40% . The majority of SCN cases are due to a germline ELANE mutation . Acquired mutations in CSF3R occur in >70% sMDS/sAML associated with SCN . Hypotheses underlying our model are: an ELANE mutation causes SCN; CSF3R mutations occur spontaneously at a low rate; in fetal life , hematopoietic stem and progenitor cells expands quickly , resulting in a high probability of several tens to several hundreds of cells with CSF3R truncation mutations; therapeutic granulocyte colony-stimulating factor ( G-CSF ) administration early in life exerts a strong selective pressure , providing mutants with a growth advantage . Applying population genetics theory , we propose a novel two-phase model of disease development from SCN to sMDS . In Phase 1 , hematopoietic tissues expand and produce tens to hundreds of stem cells with the CSF3R truncation mutation . Phase 2 occurs postnatally through adult stages with bone marrow production of granulocyte precursors and positive selection of mutants due to chronic G-CSF therapy to reverse the severe neutropenia . We predict the existence of the pool of cells with the mutated truncated receptor before G-CSF treatment begins . The model does not require increase in mutation rate under G-CSF treatment and agrees with age distribution of sMDS onset and clinical sequencing data .
Cancer development is driven by series of mutational events , which may become fixed in a hematologic or non-hematologic tumor via genetic drift . This process usually includes a limited number of driver ( advantageous ) mutations , and a greater number of passenger ( neutral or mildly deleterious ) mutations . Driver mutations for several hundred different cancers have been identified by sequencing and functional assays . The relationship between driver and passenger mutations has been investigated using mathematical models representing carcinogenesis in terms of a “tug-of war” between the former and the latter [1 , 2] . Another related problem is whether carcinogenesis is driven by acquisition of single point mutations or by saltatory changes amounting to major genome rearrangement events [3 , 4] . Mathematical modeling of interactions among multiple drivers has been described by Nowak and Durrett and their colleagues [5–7] . These frequently involve branching processes and related mathematical models [8] . Among stochastic models in hematology , an example is [9] . Hematopoiesis provide the best-characterized system for cell fate decision-making in both health and disease [10] , as well as connections between stimuli such as inflammation and cancer [11] . Here , we model a disease evolving on the background of a germline mutation . The acquired driver mutation recurs during tissue expansion phase in fetal life and becomes selectively advantageous in early childhood , leading to development of malignancy . As a prominent example of such disease , we model the important hematologic disorder Severe Congenital Neutropenia ( SCN ) , a monogenic inherited disorder , that acquires new mutations and evolves to secondary myelodysplastic syndrome ( sMDS ) or secondary acute myeloid leukemia ( sAML ) . This model is similar to the “fish” graph of Tomasetti and Vogelstein [12]; however the latter is more comprehensive and involves multiple driver case . Here , we use tools of population genetics and population dynamics to model progression from SCN to sMDS and dissect the contributions of mutation , drift and selection at different stages of an individual’s life . More specifically , we consider: Accordingly , SCN is most commonly due to germline mutations in ELANE , which encodes the neutrophil elastase [13] . SCN is characterized by the near absence of circulating neutrophils , which renders the child , typically an infant , susceptible to recurrent life-threatening infections . The introduction in the 1990s of recombinant granulocyte colony-stimulating factor ( G-CSF ) to increase circulating neutrophils , markedly improved survival and quality of life for SCN patients [14] . SCN often transforms into sMDS or sAML [15 , 16] . Clinical studies have demonstrated a strong association between exposure to G-CSF and sMDS/AML [17–21] . Mutations in the distal domain of the Granulocyte Colony-Stimulating Factor Receptor ( CSF3R ) have been isolated from almost all SCN patients who developed sMDS/AML [22 , 23] . Clonal evolution over approximately 20 years was documented using next generation sequencing and quantification of CSF3R allele frequency variation in an SCN patient who developed sMDS/sAML [24] . Strikingly , out of four different mutations in CSF3R , one persisted into the leukemic clone but the other three were lost , supporting the assumption of different selective values in the presence of G-CSF that underlies our model . As clonal evolution is a central feature in cancer [25–28] and next generation sequencing has revealed complex genomic landscapes , SCN may provide a simpler real-world example to study cancer development . Two opposing paradigms have been proposed for cell fate decision making in blood cells: stochastic hematopoiesis ( based on variability observed in cultured bone marrow cells as first suggested by McCulloch and Till [29] ) and deterministic , or instructive , hematopoiesis ( growth factor-driven production of specific blood cell types ) [30 , 31] . In spite of substantial experimental findings , particularly recent single-cell measurements [32] , the two opposing theories await a grand synthesis . Disease-accompanying dynamics have been variously modeled over the years as deterministic or stochastic[10] . SCN may also provide a simpler real-world example to study cell fate determination . Little is known about the molecular mechanism ( s ) by which SCN leads to myeloid malignancy and how important are the truncating mutations such as CSF3R D715 in this process . Notwithstanding the exact molecular mechanism by which the CSF3R truncation mutants lead to sMDS/AML , two pivotal questions concerning the population dynamics and population genetics of the mutant clones are: ( i ) whether the CSF3R truncation mutants are present before application of G-CSF , and ( ii ) whether G-CSF administration increases mutation rate in hematopoietic stem cells with ELANE mutation . If the answer to the first question is affirmative , then the presence of a small subpopulation of CSF3R truncation mutants among infants with SCN might be of prognostic value and a preventive therapy might be sought . Concerning the second question , determination whether G-CSF is mutagenic or provides a selective pressure may influence the degree G-CSF therapy is conducted , e . g . should it be more or less aggressive . To provide insight into the course of this disease and its clinical management , we propose a novel model of the emergence and fixation of CSF3R truncating mutations , which also follows the paradigm outlined earlier on . We note that the most common of these mutations associated with transition to sMDS is the CSF3R D715 . However , at the resolution level of our model , we are not able to make more specific distinctions nor to consider coexistence or competition of more than one truncation mutant . The model assumes that the answer to question ( i ) is affirmative , but it is negative to question ( ii ) . The model’s hypotheses are: ( B ) Effect of CSF3R ( wild type vs . D715 ) on cell proliferation . Ba/F3 cells expressing either wild type ( Type I ) or mutant ( D715 ) CSF3R were treated with increasing doses of recombinant human G-CSF ( ng/ml ) and proliferation was measured by the MTT assay performed in triplicates in a 96 well plate . The data are raw absorbance values at 600 nm and represent the three replicates plotted against increasing dose of G-CSF , fitted using least squares by Hill-type curves ( Type I , blue , D715 red ) . For details of experimental procedures see the S1 Appendix . Fitting and statistical procedures are explained in the Methods . We show that hypotheses 2 and 3 are needed , by first building a proof-of-principle simple Moran process ( a stochastic model used in population genetics ) with no expansion , which fits the data only if it is started by a cell population including ⁓101-102 cells expressing a CSF3R truncation mutation . Then we show that this number of CSF3R truncation mutant cells can be produced in the late fetal period expansion of hematopoiesis in bone marrow . Existence of the pool of CSF3R truncation mutant cells before exposure to G-CSF can be discovered only by deep targeted sequencing . Then we follow up with a full-fledged comprehensive model , which accounts for the important detail of time change of the size of the hematopoietic system , but which confirms the conclusions of the proof-of-principle model .
The results of the proof-of-principle modeling summarized in Table 1 suggest that it is feasible to build a more comprehensive model consistent with normal hematopoiesis as well as mutation and selection mechanisms modified by the age-dependent cellularity of the bone marrow and administration of pharmacological G-CSF .
Here we presented a model of fixation of a CSF3R truncation mutant in the transition from an inherited neutropenia to sMDS: from the expansion phase in the prenatal hematopoietic tissues , to initiation of the G-CSF treatment , to expansion of the mutant , and to replacement of the normal bone marrow by the pre-leukemic mutants . By modifying the simple Moran model of population genetics , we provided an explanation for the evolution of sMDS in about 70% of cases in which CSF3R truncation mutant acts as an oncogenic driver . We first used a proof-of-concept two-stage model including the initial creation of the mutant clone before the selective agent G-CSF has been applied , followed by the period of selective pressure after initiation of treatment . We followed up with a more comprehensive model , which used the estimates of age-dependent productivity changes in hematopoietic stem cells , obtained based on telomere shortening estimates by the Abkowitz and Aviv groups [47 , 48] . Our model provides a real-world setting that may further illuminate principles of clonal hematopoiesis of indeterminate potential , first described as age-related clonal hematopoiesis . As recently summarized by [50] , HSC clonality and association with malignancy begins with somatic genetic lesions in adult stem cells that accumulate and persist and that “given a large enough population ( of HSC ) , every base pair in the genome will be mutated within at least one HSC” . Further , “these mutations provide the substrate for clonal selection” . The original and distinctive feature of our present model is to show that mutations occurring during the bone marrow expansion in the fetal period are likely to play a major role in creating this substrate . The expected times to fixation of the CSF3R truncation mutant ( 4–22 years ) are consistent with the timing of the sMDS onset . According to data published from the European SCN Registry data , the average age at diagnosis of SCN with sMDS and CSF3R mutation is 13 ± 9 years [51] . The 70% fixation probability requires 11–60 “initial” cells harboring the mutation . We experimentally validated our mathematical model by measuring the growth advantage of the CSF3R D715-expressing cells and found a significant growth advantage ( Fig 1A ) . Further validation will require next generation sequencing of specimens from these rare patients . Qiu et al . [52] recently reported that this truncation mutation also permits granulocytic precursors to avoid apoptosis . Our comprehensive model is based on the hypothesis that the rate of cell division after birth , when the rapid expansion of bone marrow slows down , is still very high . Hence , acquisition of new mutants during that phase is still substantial . However , selection is the force that leads the mutant-receptor cells to dominate . This also means that supply of new mutants in the expansion phase might not be necessary for the disease to emerge . However , it is likely that proliferation slows down by one or two orders of magnitude , depending on exact characteristics of subtypes of stem cells . Then in order to fit the data , somewhat higher selection coefficients are needed . In that case , the comprehensive model will behave approximately as the “proof of the concept” model , i . e . most of the mutant are supplied in the marrow expansion stage . An alternative hypothesis states that an inherited neutropenia induces a maladaptive increase in replicative stress and higher mutation rate in HSC that contributes to transformation to sMDS/AML [53] . However , measurements of the mutation burden in individual hematopoietic stem/progenitor cells ( HSPCs ) from SCN patients failed to support that . CD34+CD38- cells were sorted from blood or bone marrow samples and cultured for 3–4 weeks on irradiated stromal feeder cells . The exomes of the expanded HSPC clones were sequenced with unsorted hematopoietic cells from the same patient served as a normal control . The average number of somatic mutations per exome was 3 . 6 ± 1 . 2 for SCN , compared to 3 . 9 ± 0 . 4 for the healthy controls . Those patient-derived findings support our model . Our conclusions require that the mutation rate per site per cell division equals about 10−9 , which is consistent with normal mutation rate in human genome . This latter issue warrants discussion since the somatic mutation rate in humans is about two orders of magnitude higher than the germline mutation rate , as suggested by [54] . However , a recent paper by Milholland et al . [55] , argues that this former ( somatic rate ) is of the order of 10−9 per base per mitosis , while the former ( germline rate ) is of the order of 10−11 per base per mitosis ( Fig 1B in that paper ) . Moreover , as seen in our Fig 2 , using the 10−7 mutation rate [54] would only slightly change our conclusions . Two other mechanisms drive the expansion of the CSF3R truncation mutants , ( i ) the initial CSF3R truncation mutant cell clones arising in the expansion phase of fetal hematopoietic bone marrow and ( ii ) competitive advantage of the CSF3R truncation mutant harboring cells at later ages , hypothetically due to increased G-CSF pressure . “Mutator phenotype” does not need to be invoked in the SCN progression to sMDS . A characteristic feature of human cancers is their wide heterogeneity with respect to extent of involvement , genotype , and rate of progression and spread [56] . This variability contrasts markedly to induced animal tumors , which grow at a relatively uniform rate . sMDS/AML secondary to SCN is not an exception , with onset varying from 1 to 38 years of age . Previously , we constructed a stochastic model of the SNC→sMDS→sAML transition based on stochastic events [9] . It considered each new mutation to provide more selective advantage to the arising clone . This linear structure of mutation conferred desirable simplicity to modeling but was not necessarily realistic . In the framework of multitype branching processes and special processes such as Griffiths and Pakes branching infinite allele model [57 , 58] , more complicated scenarios might be contemplated . Interestingly , the model of ref . [9] suggests that the spread in the age of onset of sAML is not due solely to stochastic nature of clone transitions , but requires a large variability in proliferative potential from one affected individual to another . Similar effect can be predicted in the Moran process . According to [38] , the time course of the Moran process under mutant selective advantage can be split into three periods: ( 1 ) relatively long period from small number of mutant cells to a threshold , followed by ( 2 ) a much shorter period from the threshold to near-fixation of the mutant , and ( 3 ) a relatively long period to complete fixation . Accordingly , once the mutant count exceeds certain threshold , the process accelerates . This results in the spread of times to fixation depending at least as strongly on the selection coefficient as on the “intrinsic” randomness . This justifies the approach we took in this study , to concentrate on the effects of the selection coefficient . In addition , determination of the threshold may help establish a target for monitoring the progress of the disease . Gaining more insight will require a further study . While our model advances the understanding of multistep progression to cancer with a real-world condition and application to the clinic , other factors could be incorporated . These include: a correlation between G-CSF dosage for neutrophil recovery in SCN patients and the risk of malignant transformation and acquisition of an additional mutation , such as RUNX1 , in the evolution to sAML . In the current report , we focus on a single aspect of the SCN-related leukemogenesis: expansion of CSF3R truncation mutant cells leading to the sMDS transformation . The model we present here provides potentially testable hypotheses ( i ) the CSF3R truncation mutants are present in 101-102 cells before G-CSF treatment is applied and ( ii ) a slight selective advantage of the CSF3R truncation mutant-harboring cells under G-CSF pressure is sufficient to lead to their expansion . The second hypothesis seems to be consistent with findings in ref [53] . Current dogma holds that clonal dynamics in relation to the development of sMDS/AML are highly heterogeneous and unpredictable . Our model supports the clinical value of more accurate disease surveillance with next generation sequencing and better timing of therapeutic interventions , such as stem cell transplantation .
|
Cancer develops by multistep acquisition of mutations in a progenitor cell and its daughter cells . Severe congenital neutropenia ( SCN ) manifests itself through an inability to produce enough granulocytes to prevent infections . SCN commonly results from a germline ELANE mutation . Large doses of the blood growth factor granulocyte colony-stimulating factor ( G-CSF ) rescue granulocyte production . However , SCN frequently transforms to a myeloid malignancy , commonly associated with a somatic mutation in CSF3R , the gene encoding the G-CSF Receptor . We built a mathematical model of evolution for CSF3R mutation starting with bone marrow expansion at the fetal development stage and continuing with postnatal competition between normal and malignant bone marrow cells . We employ tools of probability theory such as multitype branching processes and Moran models modified to account for expansion of hematopoiesis during human development . With realistic coefficients , we obtain agreement with the age range at which malignancy arises in patients . In addition , our model predicts the existence of a pool of cells with mutated CSF3R before G-CSF treatment begins . Our findings may be clinically applied to intervene more effectively and selectively in SCN patients .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion"
] |
[
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"maternal",
"health",
"obstetrics",
"and",
"gynecology",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"immunology",
"bone",
"marrow",
"cells",
"oncology",
"physiological",
"processes",
"mutation",
"women's",
"health",
"granulocytes",
"stem",
"cells",
"birth",
"white",
"blood",
"cells",
"animal",
"cells",
"carcinogenesis",
"point",
"mutation",
"hematopoietic",
"stem",
"cells",
"cell",
"biology",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"hematopoiesis"
] |
2019
|
Mutation, drift and selection in single-driver hematologic malignancy: Example of secondary myelodysplastic syndrome following treatment of inherited neutropenia
|
DNA double-strand breaks are lesions that form during metabolism , DNA replication and exposure to mutagens . When a double-strand break occurs one of a number of repair mechanisms is recruited , all of which have differing propensities for mutational events . Despite DNA repair being of crucial importance , the relative contribution of these mechanisms and their regulatory interactions remain to be fully elucidated . Understanding these mutational processes will have a profound impact on our knowledge of genomic instability , with implications across health , disease and evolution . Here we present a new method to model the combined activation of non-homologous end joining , single strand annealing and alternative end joining , following exposure to ionising radiation . We use Bayesian statistics to integrate eight biological data sets of double-strand break repair curves under varying genetic knockouts and confirm that our model is predictive by re-simulating and comparing to additional data . Analysis of the model suggests that there are at least three disjoint modes of repair , which we assign as fast , slow and intermediate . Our results show that when multiple data sets are combined , the rate for intermediate repair is variable amongst genetic knockouts . Further analysis suggests that the ratio between slow and intermediate repair depends on the presence or absence of DNA-PKcs and Ku70 , which implies that non-homologous end joining and alternative end joining are not independent . Finally , we consider the proportion of double-strand breaks within each mechanism as a time series and predict activity as a function of repair rate . We outline how our insights can be directly tested using imaging and sequencing techniques and conclude that there is evidence of variable dynamics in alternative repair pathways . Our approach is an important step towards providing a unifying theoretical framework for the dynamics of DNA repair processes .
Double-strand breaks ( DSBs ) are lesions in DNA that occur naturally by oxidative stress , DNA replication and exogenous sources [1 , 2] . When left unprocessed or during erroneous repair , they cause changes to DNA structure creating mutations and potential genomic instability [3–8] . To repair DSBs , multiple mechanisms have evolved and are known to include non homologous end joining ( NHEJ ) [7 , 9–17] , homologous recombination [18] including single strand annealing ( SSA ) [19 , 20] , microhomology mediated end joining ( MMEJ ) [21 , 22] and alternative or back-up end joining ( A-EJ ) [23 , 24] . The choice of mechanism depends on the structure of the break point , where simple breaks caused by restriction enzymes are different in structure from those caused by ionising radiation ( IR ) ( reviewed in [25 , 26] ) . This affects the probability of error prone repair because mutations are mechanism specific and depending on which mechanism is activated a cell might exhibit chromosome translocations [4 , 5] , small deletions or insertions [6 , 7] or recombination leading to loss of heterozygosity [8] . For example , in mouse , error by SSA causes chromosome translocations [4] and in Saccharomyces cerevisiae , NHEJ of simple DSBs is associated with small deletions or insertions [7] . In vivo studies of DSBs have suggested that in addition to structural activation arising from the break point , cell-cycle dynamics can also play a role in repair mechanism activation ( reviewed in [27] ) . In particular the choice of mechanism is not fixed at the time of damage and cells exhibit a pulse like repair in human U-2 OS cells [28] , a behaviour supported by a molecular basis for cell cycle dependence in NHEJ , mediated by Xlf1 phosphorylation [29] . To understand how mutations are distributed in the genome , it is important to uncover the dynamic activation and interplay between different DSB repair mechanisms . This mutual activation is not fully understood , however the individual repair mechanisms and recruitment proteins of NHEJ , SSA and A-EJ have been documented . NHEJ requires little or no homology , is a mechanism of DNA end joining in both unicellular and multicellular organisms [7] and can exhibit fast repair by the binding of DNA dependent protein kinase ( DNA-PK ) [15] . In vertebrates , NHEJ initiates the recruitment and binding of several proteins . These have been shown to include Ku70 , Ku80 , DNA-PK catalytic subunit ( DNA-PKcs ) , Artemis and Ligase IV in a cell free system [9] . Ku70 and Ku80 are subunits of the protein DNA-PK . Biochemical and genetic data suggest they bind to DNA ends and stimulate the assembly of NHEJ proteins by DNA-PKcs [10 , 12] . Repair proceeds by Artemis facilitated overhang processing and end ligation via DNA Ligase IV [13 , 14] . Although well studied , new regulating components of NHEJ are still being discovered , for example the protein PAXX [17] . SSA is slower than NHEJ and in yeast works at almost 100% efficiency for homologous regions of at least 400bp [6] . First described in mouse fibroblast cells [19 , 20] , during SSA two complementary sequences are exposed through a 5′ to 3′ exonuclease end resection and aligned . Remaining overhangs are then cut by an endonuclease and the DNA is reconstructed by DNA polymerase using the homologous sequences as a template . Some of the components that contribute to SSA have been identified in eukaryotes e . g . the complex MRN consisting of Mre11 , Rad50 and Nibrin which facilitates DNA end resection [30] . Following resection , replication protein A ( RPA ) binds to the DNA and when phosphorylated , forms a complex with Rad52 to stimulate DNA annealing [31 , 32] . Similarly to NHEJ , following gap repair , SSA is terminated with end ligation by Ligase III [33] . Data of repair kinetics for mutants defective in Rad52 show limited slow repair in comparison to wild type repair curves in gamma irradiated cells in chicken B line cells [34] , suggesting that SSA may be active in the repair of DSBs caused by IR . In yeast , it has been suggested that SSA constitutes a major role in the repair of DSBs accounting for three to four times more repairs than gene conversion during M phase [35] . One interesting finding in genetic studies is that when NHEJ is compromised , DSBs are removed by alternative mechanisms , that have adopted various names in the literature , such as MMEJ in yeast [36] and back-up NHEJ ( B-NHEJ ) in higher eukaryotes [37] that here , we collectively refer to as A-EJ [23 , 38] , ( reviewed in [37 , 39 , 40] ) . It is still unclear how A-EJ is regulated or interacts with other processes but there is evidence it is active in the repair of breaks with microhomology of 3–16 nucleotides , reviewed by Decottignies [40] . Thought to act on break points with ends that are not complementary in the absence of NHEJ factors [38] , an assortment of PARP-1 , 53BP1 , Lig3 and 1 , Mre11 , CtIP and Polθ have been proposed as regulators of A-EJ . PARP-1 is required and competes with Ku for binding to DNA ends through the PARP-1 DNA binding domain [24] . Other proteins are involved in initial binding , where activation of 53BP1 in MMEJ is dependent on Ku70 and independent of DNA-PKcs [22] and CtIP has been associated through the use of microhomology [41] . The proteins required for end joining have been identified as Lig3 and Lig1 in the absence of XRCC1 [42–44] . This pathway has never been observed in single cells and it is unclear how A-EJ is related to other mechanisms . However , targeted RNAi screening for A-EJ has uncovered shared DNA damage response factors with homologous recombination [45] . For an illustration of the three mechanisms ( see Fig 1a ) . Mathematical models of DSB repair have used biphasic [46] , biochemical kinetic [47–52] , multi-scale [53 , 54] , and stochastic methods [55] . In a study by Cucinotta et al . [48] , a set of coupled nonlinear ordinary differential equations were developed . The model was based on the law of mass action and stepwise irreversible binding of repair proteins to describe NHEJ rejoining kinetics and the phosphorylation of H2AX by DNA-PKcs . Similar studies have modelled repair kinetics and protein recruitment during SSA [49 , 51] , NHEJ [50] and other mechanisms including NHEJ , HR , SSA and two alternative pathways under a wide range of linear energy transfer ( LET ) values and heavy ions [52] . These studies have been met with some controversy , for example with the argument that the biphasic model has never succeeded in providing definitive values for the repair components [56] . Recently however , the models have been further developed to model the complexity of a DSB by application to damage induced by ionising radiation of different qualities [57 , 58] . This can be achieved because the spectrum of DSB-DNA damage can be computed by applying Monte Carlo track structure [59] , which is a method that can be used to simulate the passage of charged particles in water , for a review see [60] . Previous biochemical kinetic models have been used to reproduce the experimental data observed . This approach often uses more parameters than are required to describe sequential steps in the repair process . This can cause difficulty in identifying parameter values because multiple parameter value combinations may be able to describe the data well , an issue known as non-identifiability . Consequently , predictions are not unique , which can be detrimental in the design of a biological experiment . Therefore the creation of models that provide a unique interpretation of repair dynamics is a challenge . Here we develop a statistical model that can take DSB repair curve data , such as those generated from pulse field gel electrophoresis ( PFGE ) or comet assays , and infer repair mechanism activation . The method relies on training a simple model against multiple data sets of DSB repair under different genetic knockouts when multiple repair mechanisms are activated . Using the most probable set of parameter values , we can then simulate the model and make predictions on the activation of different rates of repair . Unlike previous modelling approaches , we do not model individual recruitment proteins . Instead we assign parameter values to different rates of repair . This has two benefits . Firstly it provides a method to uncover different rates of repair arising from different repair mechanisms that are implicit in the data . Secondly , it reduces the number of parameters required to describe the system , leading to a more identifiable model . Our approach strikes a balance between a detailed mechanistic description of the biochemical components with a traditional statistical model . This enables insights into the dynamical process underling repair pathways combined with novel and testable predictions . We use this method to integrate the data from eight repair curve assays under genetic knockouts including combinations of Ku70 , DNA-PKcs , Rad52 and Rad54 . We first infer that there are at least three disjoint dynamical repair mechanisms that explain the combined data and that the dynamics depend on the regulating recruitment proteins . We propose that there are a number of alternative end joining dynamical processes that may or may not share a common genetic pathway . We also demonstrate that our model has predictive power on new data sets , including PARP-1 knockouts , and show that the activation of different repair processes over time depends on the speed of the underlying dynamics .
The experimental data used in this study are published repair curves generated from methods of pulse field gel electrophoresis , a technique that distributes the DNA according to the length of the fragment . We model the dose equivalent number of DSBs that are obtained from the fraction of DNA released into the gel [61] . Table 1 lists the experimental data that are used for inference . Cells were exposed to X-rays [24] and the number of DSBs within the population recorded over time . The eight data sets are labelled D1–D8 . Data D1 is wild type and since the cell cycle phase is unrestricted , we expect all three repair processes to be present . Data D2 and D3 are DNA-PKcs knockouts in G1 and G2 phase , where we expect NHEJ to be compromised but since Ku is present we still expect the recruitment process . Data D4 is a Rad52 knockout where we expect only NHEJ and A-EJ to be present . Data D5 and D6 are Ku knockouts , where we assume the whole of the NHEJ pathway to be compromised and only SSA and A-EJ remain active . Data D7 and D8 are expected to have no repair by PARP-1 mediated A-EJ because both sets were treated with PARP-1 inhibitors . Data D7 comes from Ku70−/− mouse fibroblasts , where we expect to see no repair by NHEJ as well as a lack of A-EJ due to PARP-1 inhibition [24] . We assume DSBs caused by ionising radiation ( IR ) can be repaired by multiple processes with different rates corresponding to NHEJ , SSA and A-EJ ( see Fig 1b ) . In principle , other mechanisms could be included , such as Rad54 dependent homologous recombination ( HR ) , but since this mechanism is mostly active in S phase of the cell cycle , and thought to contribute little to IR induced DSBs [62] , we chose not to model it explicitly . We model a DNA repair process by a stochastic reaction system , represented by x + E i → K i y i ( 1 ) y i → K i ′ ∅ + E i , ( 2 ) where x is the DSB , Ei is the recruitment protein for process i ( Ku , MRN and PARP-1 ) , and Ki , K i ′ are the parameters for recruitment and subsequent ligation respectively . To model the limited resources available to the cell , we follow the approach of Cucinotta et al [48] , and assume that the total amount of protein is conserved for each repair mechanism E i + y i = C i , ( 3 ) where Ci is the total amount of recruitment protein for each repair mechanism . This structure also captures the experimentally observed fact that mRNA levels for repair processes depend on the radiation dose [64] . The reactions result in a nonlinear coupled stochastic system which is simulated using the Gillespie algorithm [65] . The proportions of DSBs repaired by each mechanism are estimated by calculating the cumulative number of DSBs that enter each individual pathway with the integral N ˜ i = ∫ 0 ∞ K i x E i d t . ( 4 ) This integral can then be used to calculate the proportion of DSBs repaired by each mechanism . To build a model that can be used to obtain unique predictions , it is advantageous to minimise the number of parameters that describe the system . To do this , we developed a hierarchical model where the individual parameter values , K , K′ , are lognormally distributed with a common mean , μ , across all the data sets in which they are included ( see Fig 2a ) . For data sets in which a repair protein is repressed downstream of the initial protein that binds , we impose an additional hyperparameter . We include this additional hyperparameter because it is not clear if a repair mechanism remains active when individual regulating proteins are repressed . More formally , we wish to obtain K i d , K i ′ d for i ∈ {1 , 2 , 3 , 4} and eight data sets d ∈ {1 , … , 8} . These parameters are constrained by the repair dependent hyperparameters , γ = {μi , σ2} , where μi represents the mean of a lognormal distribution , and σ2 the variance . The K i d , K i ′ d are drawn from the population level distributions , where the joint density can be written π ( D , K , γ ) = π ( D | K ) π ( K | γ ) π ( γ ) , ( 5 ) and Bayes rule becomes π ( γ | D ) = π ( γ ) ∫ f ( D | K ) π ( K | γ ) d K π ( D ) , ( 6 ) which gives the posterior of the hyperparameters given the data , D . The integral indicates that we sum over ( marginalise ) the K values . To perform the inference we use approximate Bayesian computation sequential Monte Carlo ( ABC SMC ) [66–69] . This method that can be used to fit a model to multiple data sets when the likelihood is unavailable . In the Bayesian framework , we are interested in the posterior distribution πϵ ( θ , x|y ) , where θ is a vector of parameters and x|y is the simulated data conditioned on the experimental data . To obtain samples from the posterior distribution we must condition on the data y and this is done via an indicator function I A y , ϵ ( x ) . We then have π ϵ ( θ , x | y ) = π ( θ ) f ( x | θ ) I A y , ϵ ( x ) ∫ A y , ϵ × Θ π ( θ ) f ( x | θ ) d x d θ , where A y , ϵ = { x ∈ D : ρ ( x , y ) ≤ ϵ } , ρ : D × D → R + is a distance function comparing the simulated data to the observed data and πϵ is an approximation to the true posterior distribution . This approximation is obtained via a sequential importance sampling algorithm that repetitively samples from the parameter space until ϵ is small , such that the resulting approximate posterior , πϵ , is close to the true posterior . The ABC SMC algorithm is used to calculate an approximation to the target posterior density πϵ ( γ|D ) , where D = {Di; i=1 , 2 , . . 8} . We can include the hierarchical model by simulating data , D* , using the following scheme: γ ∼ U ( α , β ) K ∼ L N ( μ , σ 2 ) D * ∼ f ( D | K ) , ( 7 ) where f ( D|K ) is the data generating model ( the solution to the reaction system represented by Eqs 1–3 ) . The α , β are the lower and upper limits of the uniform prior on the hyperparameters . Note that in the sequential importance sampling step , we perturb only the hyperparameters . For further details on the inference and all prior values see S1 Text .
We fit three different models , comprising one , two and three repair processes respectively ( M1: {i = 1} , M2: {i = 1 , 2} and M3: {i = 1 , 2 , 3} ) , and found that a three process model describes the best fit using an approximation to the Deviance Information Criterion ( DIC ) [70] and the Akaike information criterion ( AIC ) [71] , based on a surrogate likelihood approach [72] ( Fig B , C in S1 Text ) . The final model structure is presented in Table 2 , ( for prior distributions see S1 Text ) , and a summary of the fitted parameter values is given in Table 3 . The fit of the simulation to the data for all eight data sets is shown in Fig 2g . The fits capture the essential aspects of the repair curves and most points are consistent with the posterior median and credible regions . When we qualitatively compare this fit to that obtained by the one and two process models , M1 and M2 , we find a poor fit for M2 in data sets D2–4 and generally large credible regions for M1 ( see Fig D , E , F in S1 Text ) . The posterior distributions of the hyperparameters are shown in Fig 2b . Inspection of the interquartile range of the hyperparameters confirms that a combination of fast , slow and intermediate repair is sufficient to describe the wild type and mutant data , furthermore a two sided Kolmogorov Smirnov test between the posterior distributions for the hyperparameters confirmed that the four distributions were significantly different to one another ( μ1 , μ2 D = 1 , μ1 , μ3 D = 1 , μ2 , μ3 D = 0 . 998 , μ2 , μ4 D = 0 . 686 , μ4 , μ3 D = 0 . 752 , μ4 , μ1 D = 1 , all tests p < 2 . 2e−16 ) . For each data set ( D1–D8 ) the posterior interquartile ranges of the parameters K1 , K2 and K3 were recorded ( Fig G in S1 Text ) . Marginal distributions for the wild type K1 , K2 , K3 are shown in Fig 2c–2e ) . Analysis of the marginal distribution shows that the parameter distributions of K1 , K2 and K3 deviate from the hyperparameter distributions , suggesting that although the rates are defined as fast , slow and intermediate , there is variation in activation of the mechanisms among different mutants ( Fig 2c–2e ) . There is some overlap in parameter values K1 , K2 and K3 ( Fig 2f ) but the interquartile ranges of the parameters K1 , K2 and K3 are disjoint , this is also observed in all eight data sets ( Fig G in S1 Text ) . For all posterior distributions of the parameters and a plot of individual DSBs and their repair in a wild type model , see Fig H-K in S1 Text . When individual DSBs are tracked in the model , the DSBs are quickly distributed amongst the three pathways and repaired according to the predicted rate . To check that our model parameters were robust to adding additional data , we performed ABC SMC again on nine data sets . This new data set consisted of the eight data sets listed in Table 1 and an additional repair time series of xrs-6 cells deficient in Ku80 inhibited of PARP-1 by DPQ [24] , in which cells were deficient in NHEJ and A-EJ . The results of the total number of DSBs repaired and our predictions on the activation of A-EJ using this data set were the same and a comparison between the eight data and nine data posterior showed a similar fit ( Figs L and M in S1 Text ) . In summary , we conclude that a three process model provides the best fit to the data observed ( Fig B-C in S1 Text ) and that for equal prior ranges on the processes the biological data can be explained by one fast , one slow and at least one intermediate rate of repair . By re-simulating from our fitted posterior distribution we were able to examine the dynamics of DSB repair across mechanisms and data sets ( see Fig 3 ) . Data sets in which NHEJ is active exhibited a faster repair with the cumulative number of DSBs reaching to within 80% of the total within a period of 2 hours post irradiation . Next , we plotted the number of DSBs entering each repair mechanism as a time series ( Fig 3b and 3d ) . The simulated data predicts that fast repair consistently processes most of the DSBs within two hours after radiation ( red curves in Fig 3 ) . Similarly , there were no clear differences amongst the data in the DSB processing by slow repair . Intriguingly , intermediate repair was slower in cells compromised of Ku70 ( D5 , D6 ) than those without DNA-PKcs ( D2 , D3 ) ( green curves , Fig 3b and 3d ) . To calculate the predicted number of DSBs repaired by fast , slow and alternative mechanisms , we computed the integral N ˜ i . The results are shown in Fig 4 . Data sets for which cells were deficient in regulating components of NHEJ confirmed variation in the numbers of DSBs repaired by intermediate rates . In agreement with the results obtained from the time series plots ( Fig 3b and 3d ) there was a difference in the ratio of slow and intermediate rates between data sets D2 , D3 and D5 , D6 . We also observed an increase in the number of DSBs repaired by A-EJ between G1 and G2 ( Fig 4 D2 , D3 ) , agreeing with experimental results in the literature [62] . To test the predictive capability of the model , we re-simulated repair curves using the posterior parameters for data sets D1–D8 and compared the simulated curves to new data sets from the literature ( Fig 5 ) . The model trajectories provide a good fit to wild type data at 40Gy ( Fig 5a , [62] ) . More impressively , the simulations of cells defective in PARP-1 by application of 3’-AB and Ku80 deficient xrs-6 cells exposed to DPQ defective in PARP-1 and NHEJ show good agreement with the experimental curves ( Fig 5b and 5c , [24] ) . This analysis suggests that small ranges of the rate parameters for the fast and slow repair processes , when assigned to NHEJ and SSA , can predict multiple low LET data sets under different experimental conditions . This suggests that they are constrained across experimental and biological conditions . Next we investigated whether the intermediate rates from our model could potentially fit the experimental data when activation of fast and slow rates are inhibited . We re-simulated the model again but this time set the parameters for active NHEJ and SSA to zero across the models for D1–D8 . The results are shown in Fig 5d along with the corresponding experimental data . In wild type data , we see that it is possible for all the DSBs to be removed with the predicted rates of A-EJ , although repair is slower , suggesting that the ability for A-EJ to repair DSBs is not saturated . When Rad52 is inhibited ( data D4 ) , because PARP-1 could still compete with MRN we see that it is not possible with the predicted rates for all the DSBs to be repaired in the absence of NHEJ . There was no clear difference in the repair of DNA-PKcs mutants , however when Ku70 is inhibited ( data D5 , D6 ) we see that when competition by slow repair is removed the predicted rates of A-EJ , suggest that the activation of A-EJ is not saturated and is perhaps inhibited by competition with MRN . The time taken for over half the DSBs to be repaired by A-EJ is shown in Fig 6a . Some repair is fast , occurring within two hours , however , for cells deficient in Ku70 , A-EJ adopts a slower repair with half maximum achieved at eight hours . The activation of A-EJ across the data sets is represented by the interquartile ranges of the posterior distributions for K3 and K 3 ′ ( Fig 6b ) . The rate for ligation , K 3 ′ , is low in data sets D5 and D6 , suggesting that intermediate mechanisms are less active in the absence of Ku70 . The rate is highest in G2 when DNA-PKcs is inhibited . These data suggest that A-EJ adopts a slow or fast repair and that the speed of repair depends on the presence or absence of DNA-PKcs and Ku70 , because inhibition of Rad52 had little effect on the time until half-maximum . There are two ways in which this difference between Ku70 and DNA-PKcs mutants can be interpreted . The first is that when Ku70 is inhibited , two alternative mechanisms are activated , one that is fast and one that is slow . The other interpretation is that A-EJ is one repair mechanism that repairs at a slower rate when Ku70 is inhibited . We also modelled the complete inhibition of A-EJ by active NHEJ ( A-EJ removed from data sets D1 and D4 ) and found that the model still captures all the observed dynamics , albeit with the rate of fast repair slightly increased in data set D4 where there is inhibition of Rad52 ( see Fig N in S1 Text ) . Inspection of the time series data ( Fig 3b and 3d ) suggests that at time t = 0 . 5hrs , the majority of DSBs that are being processed are within a fast mode of repair . Fig 6c illustrates a typical distribution of DSBs over each mechanism at different points in time . At time t = 0 , the cells are exposed to a single dose of ionising radiation . Quickly , for example at time t < 1hrs , fast and possibly faster alternative repair mechanisms dominate the DSB processing . Later , after all DSBs processed by the faster mechanisms have been repaired , the remaining DSBs fall within the category of breaks that require processing by slower mechanisms . This change in the activity of repair mechanisms could potentially be investigated by recording changes in the level of recruitment proteins or gene expression as time series . To quantify this change in our simulated data , we plotted the percentage of DSBs that remain in active repair mechanisms over time for the wild type data ( see Fig 6d ) . By inspection , we can see that at 0 . 5 hours after irradiation most DSBs reside in the fast mechanisms . At a time of t = 8hrs , the percentage approches zero for fast repair and the majority of DSBs are found within the slow repair pathways . The variation in Fig 6d is shown with the 25th and 75th percentiles and this is due to the variation in repair rate K i , K i ′ for each mechanism . Ultimately , it is the values of the parameters K i , K i ′ that determine the rate of repair , so to confirm if the dynamics presented in Fig 6c are representative of the whole data set , we considered all time series for parameters Ki , a total of 9000 simulations . For each parameter at every time point , we assigned a value of 1 if the corresponding mechanism for the parameter contained over 30% of the total DSBs being processed at that time point and a value of 0 if it contained less than 30% . The results are shown in Fig 6e , where for each parameter Ki , a red line indicates the times at which the mechanism with rate Ki is greater than 30% active . There is a clear trend showing that the percentage of total activation decreases in time with an increase in repair rate K . In other words , the model predicts the times at which different repair processes are likely to express regulating components . When repair is extremely slow the repair mechanism never reaches 30% of the current DSBs . In summary , these results predict that if a cell experiences a sudden creation of DSBs , then gene expression for slower repair mechanisms will be maintained for longer than those required for faster repair mechanisms such as NHEJ , a result that has been implied for NHEJ and HR ( Fig 3 in [28] ) .
In this study , we presented a new hierarchical model of DSB repair and applied Bayesian inference to infer the number of active repair processes and their dynamical behaviour from experimental PFGE data . Because the model assumptions are simple and exclude the full mechanistic details of the biological processes , we are able to form an identifiable model and provide unique insights on the difference in dynamics under multiple knock-out cell lines . We have identified four major insights , each of which can be further tested experimentally . The first insight is that the data is explained by at least three independent mechanisms . Our results suggest that there are multiple dynamic regimes for the intermediate process . For example a mechanism faster than Rad52 dependent HR is required to fit the experimental data to the model in data sets D2 and D3 ( knockout of DNA-PKcs ) . Another interesting finding is that intermediate repair is increased in G2 phase of the cell cycle . If we assume that intermediate repair corresponds to alternative end joining , then this is in agreement with experimental results in the literature , supporting the existing biological evidence of the role of A-EJ in DSB repair [62] . This agrees with genetic studies that suggest two forms of alternative end joining depending on the presence of microhomology [40] . Our second insight is that the speed of A-EJ depends on the presence of regulating components in NHEJ and SSA , and in particular we observe a slower rate when Ku70 is inhibited and a faster rate when DNA-PKcs is inhibited ( data sets D5–D6 and D2–D3 respectively , Fig 3 ) . As reported by experimental analysis , Ku deficient cells do not produce NHEJ products due to excessive degradation or inhibition of end joining [11] , and our model suggests that a slower A-EJ is active in these cells ( data sets D5–D6 ) . In addition , inhibition of DNA-PKcs does not activate repair by PARP-1 mediated A-EJ [24] and leads to elevated levels of resection and more HR [73] . Together with our model , this suggests that an alternative mechanism that is faster than PARP-1 mediated A-EJ could be activated when DNA-PKcs is inhibited ( data sets D2–D3 ) . These results could be tested by examining DSB repair in single cells with and without inhibitors using time-lapse microscopy and existing markers such as fluorescently tagged 53BP1 , a protein that co-localises with DSBs , and fluorescent tagged PARP1 , a candidate protein for A-EJ [28 , 74] . The third insight that is generated from our analysis is the prediction of the total number of DSBs repaired by each mechanism . We can use this to estimate the proportion of different mutations following DSB repair in wildtype and mutant cells . Some cancers are deficient in at least one repair mechanism and in these cases , alternative mechanisms of repair have been observed to compensate [75] . One example is the increase in chromosomal aberrations observed in cells compromised of NHEJ by loss of Ku80 [3] . Recently , mutations specific to alternative mechanisms have been identified , where next generation sequencing has revealed sequence specific chromosome translocations following A-EJ at dysfunctional telomeres [5] . In addition , A-EJ is error prone , giving rise to chromosome translocations , of which there are more when NHEJ is inactive , suggesting it’s role as a back up mechanism in eukaryotes [44] . If we know how many DSBs are likely to be repaired by each mechanism , this information will be important in predicting the numbers and types of mutations that we expect to observe . Potentially , a better understanding of the interplay between DSB repair mechanisms could be applied to design synthetic lethal therapeutics in cancer [76] . The fourth insight is that the gene expression profile of the proteins within different DSB repair mechanisms should change over time , with slower repair mechanisms still remaining active many hours after the initial dose of radiation . Pulse-like behaviour has been recorded in the repair of DSBs in human cells [28] and we suggest that this prediction could be further investigated using microarrays or sequencing . In fact , the expression of repair pathway genes has recently been used to diagnose the prognosis of carcinomas [77] . Currently the genes involved in the different repair pathways—and how much they are shared—remains to be fully elucidated . Our model could be used to predict the times at which different repair pathways dominate and provide a theoretical model for the interpretation of time-course gene expression results . The model will require further development before it can be applied to more general problems in the DNA damage response . With additional data it will be possible to extend the model and include more terms such as explicit repressive cross-talk interactions . It may be possible that inhibition of a certain protein may not completely ablate the function of a repair pathway . While our two-step model somewhat accounts for this , the actual pathway contains many different proteins and more complex effects could arise . Additionally , the PFGE assay provides a population average of the total repair; in order to obtain information on relatively small numbers of double strand breaks , together with estimates of cellular heterogeneity and stochasticity , data obtained by methods such as live cell imaging are required . Another current limitation arises from the heterogeneity of the DNA damage spectrum , where complex breaks can account for 30–40% of the population [78] . Previous methods of Monte Carlo track structure have been able to predict the number of single strand breaks , simple and complex DSBs that are created [59] . By identifying proteins responsible for the processing of complex DSBs and analysing knockout repair curves , it will be possible to gain understanding on the repair of heterogenous breaks , and incorporate these into our model . Despite our demonstration that the model is predictive across multiple doses in the low LET regime , we modelled repair following a variety of radiation doses , and we can’t rule out that complexity of DSBs further alters the dynamics of repair . Previous studies have also modelled the formation of foci by summing the DSB enzymes involved in the phosphorylation of H2AX [48 , 52] and using this approach , our model could be developed to take into account lower doses of radiation or the dynamics of γ-H2AX foci formation . Related complicating factors are clustered DSBs [79] and DSBs residing in heterochromatic regions , which have been shown to require Artemis for repair [80] . From our simple assumptions we have built a predictive model and generated in silico data that was used to produce a number of unique insights that can be tested experimentally . Mathematical modelling not only facilitates the analysis of disparate data sets but also enforces the explicit formalization of the underlying assumptions of our hypotheses . Our framework is another step towards a theoretical understanding of the dynamics of DNA repair pathways . The DNA damage response is comprised of a large number of extremely complex interacting biological pathways . As the collection of larger and more heterogeneous data sets increases , we anticipate that mathematical modelling approaches will be absolutely essential for the reverse engineering and understanding of these complex biological processes .
|
DNA double-strand breaks occur during metabolism , DNA replication and by exposure to exogenous sources such as ionising radiation . When the genome is inflicted with this type of damage , DNA repair machinery is promoted to restore genome structure . The efficient interplay between DNA damage and repair is crucial to genome stability because the choice of repair mechanism directly affects the probability of mutation . Multiple mechanisms of DNA repair are known to exist , however , the subtleties of how they are activated and their interactions are yet to be fully determined . We hypothesise that a combination of Bayesian statistics and mathematical modeling is essential to elucidate the network dynamics . Studies in the literature have presented time series data of double-strand break repair in wild type and mutant cells . By combining existing time series data , our modeling approach can quantify the differences in activation amongst mutants and in addition identify a number of novel insights into the dynamics of the competing mechanisms . We conclude that alternative mechanisms of repair exhibit variable dynamics dependent on the levels of individual recruitment proteins of the predominant repair pathways .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"ionizing",
"radiation",
"dose",
"prediction",
"methods",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"dna-binding",
"proteins",
"radiation",
"simulation",
"and",
"modeling",
"dna",
"damage",
"pharmaceutics",
"dna",
"research",
"and",
"analysis",
"methods",
"proteins",
"physics",
"biochemistry",
"cell",
"biology",
"nucleic",
"acids",
"nuclear",
"physics",
"genetics",
"biology",
"and",
"life",
"sciences",
"dna",
"repair",
"physical",
"sciences",
"non-homologous",
"end",
"joining"
] |
2016
|
Mechanistic Modelling and Bayesian Inference Elucidates the Variable Dynamics of Double-Strand Break Repair
|
The vast majority of genome-wide association study ( GWAS ) findings reported to date are from populations with European Ancestry ( EA ) , and it is not yet clear how broadly the genetic associations described will generalize to populations of diverse ancestry . The Population Architecture Using Genomics and Epidemiology ( PAGE ) study is a consortium of multi-ancestry , population-based studies formed with the objective of refining our understanding of the genetic architecture of common traits emerging from GWAS . In the present analysis of five common diseases and traits , including body mass index , type 2 diabetes , and lipid levels , we compare direction and magnitude of effects for GWAS-identified variants in multiple non-EA populations against EA findings . We demonstrate that , in all populations analyzed , a significant majority of GWAS-identified variants have allelic associations in the same direction as in EA , with none showing a statistically significant effect in the opposite direction , after adjustment for multiple testing . However , 25% of tagSNPs identified in EA GWAS have significantly different effect sizes in at least one non-EA population , and these differential effects were most frequent in African Americans where all differential effects were diluted toward the null . We demonstrate that differential LD between tagSNPs and functional variants within populations contributes significantly to dilute effect sizes in this population . Although most variants identified from GWAS in EA populations generalize to all non-EA populations assessed , genetic models derived from GWAS findings in EA may generate spurious results in non-EA populations due to differential effect sizes . Regardless of the origin of the differential effects , caution should be exercised in applying any genetic risk prediction model based on tagSNPs outside of the ancestry group in which it was derived . Models based directly on functional variation may generalize more robustly , but the identification of functional variants remains challenging .
In the past six years , genome-wide association studies ( GWAS ) have revealed thousands of common polymorphisms ( tagSNPs ) associated with a wide variety of traits and diseases , particularly as study sample sizes have increased from thousands to hundreds of thousands of subjects . Typically GWAS analyses stratify on genetic ancestry , because many polymorphism allele frequencies differ by ancestral group , easily producing false positive associations for traits that also correlate with genetic ancestry . The large majority of GWAS results reported to date derive from analyses in populations of European ancestry ( EA ) [1] , [2] . Although GWAS in Asian populations in particular are becoming more common [3]–[6] , it remains important to understand the degree to which the magnitude and direction of allelic effects generalize across diverse populations [7]–[10] . The multi-ethnic PAGE consortium [11] provides a unique opportunity to assess GWAS generalization across multiple non-EA populations and multiple traits .
Subject and genotyping panel selection for the PAGE consortium have been described elsewhere [11] , [12] . In brief , a panel of 68 common polymorphisms previously reported to associate with body mass index ( BMI ) [13] , type 2 diabetes ( T2D ) [14] , or lipid levels [15] was genotyped in up to 14 , 492 self-reported African Americans ( AA ) , 8 , 202 Hispanic Americans ( HA ) , 5 , 425 Asian Americans ( AS ) , 6 , 186 Native Americans ( NA ) , 1 , 801 Pacific Islanders ( PI ) , and 37 , 061 EA ( for details , see Materials and Methods , Table S1 and Table S2 ) . We also analyzed a subset of 5863 AA from PAGE who were genotyped on the Illumina Metabochip , which contains approximately 200 , 000 SNPs densely focused on 257 regions with reported GWAS associations to traits that include lipids , BMI , and T2D [16] . For a replication analysis it would be overly conservative to use the Bonferroni correction , so the Benjamini-Hochberg method [17] was applied to assess replication of previous EA reports in the PAGE EA population . Reported effects in EA were replicated for 51 out of the 68 index SNPs at a 5% FDR . Power to replicate at most of these 68 SNPs far exceeded 80%; 16 of the 17 SNPs that did not replicate exceeded 80% power to replicate the reported effect size , and the 17th exceeded 70% power , as described previously [13]–[15] . The originally reported effect sizes tend to be less extreme for these seventeen index SNPs , but in 63 out of 79 comparisons between non-EA and EA populations involving these 17 SNPs , the direction of effect was the same in EA and non-EA groups ( p<10−5 for the null hypothesis of random effects in either direction , data in Table 1 column “Index SNPs Not Replicated in EA” ) . Only 79 of the 85 possible pairwise comparisons against EA were assessed , because some of the 17 SNPs were not genotyped in all five non-EA populations . Thus , it appears likely that most of the 17 failures to replicate represent weak effects that were underpowered in PAGE EA , rather than false-positive primary reports . Therefore , all 68 index SNPs were carried forward in the generalization analysis . In all non-EA groups , we observe significantly more effects in the same direction as in EA than expected under the null hypothesis , ranging from 68% in Asians to 88% in Hispanics ( p<0 . 001 in all non-EA groups , Table 1 and Figure 1 ) . Even in the relatively small Pacific Islander population ( N = 1801 ) , where only four index SNPs were significantly associated with reported traits , 48 out of 62 effects were in the same direction as EA ( p<0 . 001 ) , so in larger samples from this population we would expect additional loci to generalize . Although a higher proportion of effects in the opposite direction of EA was observed in Asians and Pacific Islanders , the opposite effects were neither significantly different from no effect , nor significantly different from the observed effect in the EA population . This suggests that the greater number of effects in the opposite direction observed in these smallest groups simply reflects greater uncertainty in estimating effect sizes for these populations , rather than any true trend toward opposite effects . The proportion of effects in the same direction as EA was similar across all non-EA populations , suggesting that for at least 70% of index SNPs , a significant effect in a consistent direction will ultimately be observed in non-EA populations of adequate size . Whereas the direction of effect was consistent between EA and non-EA populations , the magnitude of effect varied considerably , consistent with prior meta-analyses of generalization [18] . Because effect sizes were correlated among non-EA populations , we applied the Benjamini-Hochberg method within each population to identify index SNPs with significantly inconsistent effects between EA and non-EA populations . Inconsistent effects ( βpop≠βEA at 5% FDR ) were observed for 17 of 68 index SNPs in at least one non-EA population ( Table 2 and Table S2 , see Box 1 for definitions ) . Inconsistent effects were most frequent in the AA population ( 12 out of 68 loci ) , but examples were also observed in Pacific Islanders and Native Americans . Although most effects were consistent between EA and non-EA populations , the relatively high frequency with which differential effects were observed in non-EA populations suggests that genetic risk models derived from GWAS in EA will predict risk less reliably in non-EA populations , particularly AA . Consequently , caution should be exercised in applying risk models based upon risk variants genotyped outside of the ethnic background in which they were derived [19] , regardless of the factors causing the observed variation between populations , . Four index SNPs showed differentially generalized effects ( ßpop≠ßEA and ßpop≠0 ) . Two of these did not replicate in EA ( rs7578597 and rs7961581 for T2D in NA ) so consistency of direction cannot accurately be inferred . Direction of effect in EA and non-EA was the same for the remaining two index SNPs; rs3764261 was significantly weaker for HDL in AA , and rs28927680 was significantly stronger for TG in Pacific Islanders . There were no observations of opposite effects where both the EA effect and the non-EA effect were significant . Considering only the 15 SNPs with a significantly inconsistent effect between EA and at least one non-EA population , 14 of 15 diluted toward the null ( p<0 . 01 , Table 2 ) , a trend driven by the AA population , where all 12 out of 12 significant inconsistencies were diluted . Expanding analysis to all 51 loci replicated in EA , regardless of whether a significant difference was observed between EA and non-EA at a given SNP , we observed a significant excess of effects diluted toward the null ( ßpop/ßEA<1 ) in AA , HA , and NA populations ( Table S5 ) . Comparisons between non-EA populations revealed that diluted effect sizes were significantly more likely in AA than in any other non-EA population . Given that differential effect sizes were observed for many tagSNPs , we sought to leverage the data in order to assess the relative contributions of several factors that might contribute to the significant trend toward diluted effects , including gene–environment interaction with an exposure that varies across populations ( differential environment ) , differences in the correlation between the index SNP and the functional variant across populations ( differential tagging ) , modulation of the index SNP effect by additional , population-specific polymorphism ( differential genetic background ) , population-specific synthetic alleles ( combinations of rare , functional alleles tagged by a single common tagSNP [20] ) , or some combination of these factors . It seems unlikely that differential environments would be much more frequent in AA than other non-EA populations , or that differential environment would consistently bias toward the null within AA . Differential tagging is consistent with differentially diluted effects in AA; because linkage disequilibrium extends over significantly shorter distances in African populations than in non-African populations [21] , [22] , common functional variants ( or synthetic alleles ) are likely to be less strongly tagged by the index tagSNPs in AA . Differential genetic background effects in AA would also be consistent with the high nucleotide diversity known to exist in this population . The rare functional variants contributing to synthetic alleles will tend to be younger than common variants , and therefore are more likely to be population-specific , so synthetic alleles are compatible with the trend toward dilution . Thus , although differential environmental effects cannot be excluded , the observed data are more consistent with differential tagging and/or differential genetic background effects , and synthetic alleles cannot be excluded . Genetic background effects can be subdivided into modifying effects , where variants elsewhere in the genome directly alter the effect associated with a given index SNP , and interference effects , where secondary variants change the proportion of variance explained by the index SNP . Interfering functional variants with effects in the same direction as the index SNP would tend to dilute the apparent effect size at the index variant . The most likely source of such variants is the region surrounding an index SNP , as demonstrably functional variants already exist in that region . Although examples have been described of genes carrying both risk and protective mutations [23]–[25] , others clearly exhibit trends toward risk alleles with similar effects ( e . g . , preferentially toward breast cancer risk alleles at BRCA1 [26] ) . If the direction of effect for functional variants in a given region is consistently biased , then an increase in the number of interfering variants within a given population would be consistent with a trend toward dilution of index effects . The higher nucleotide diversity observed in African populations relative to non-African populations [27] , [28] would be consistent with a greater burden of secondary functional variants in AA than other populations . In order to assess contribution of the factors outlined above to differential effect sizes between EA and AA in the index tagSNP associations , high density genotype data were collected from a subset of the PAGE African American sample . The number of AA individuals used for index tagSNP analyses varied by phenotype , with an average of 7501 ( Table S3 ) . Similar data on other populations are currently unavailable , so only loci showing differential effects between EA and AA could be analyzed . Genotype data were collected using the Metabochip , a high density genotyping array commercially available from Illumina . Detailed methods for the Metabochip genotype data collection , calling , and quality control are available elsewhere [12] . In order to measure the contribution of differential LD to dilution , we need a model of how changes in LD between tagSNP and a functional variant would be expected to alter the observed effect size at the tagSNP , assuming that the effect size at the functional variant is the same in both populations . Given a functional SNP ( fSNP ) and an associated tagSNP , linkage disequilibrium between the two SNPs can be described as the measurement error introduced by genotyping the tagSNP , rather than genotyping the fSNP directly . As such , by appealing to prior work on regression dilution bias , it can be shown that the effect size β′ at the tagSNP is related to the effect size β at the fSNP by the following equation: ( see Text S1 for details ) . Thus , assuming that the effect size at the fSNP is constant between populations , when linkage disequilibrium between tagSNP and fSNP is weaker in a given population , we expect to see a greater degree of dilution bias for the estimated tagSNP effect size . Rearranging this equation , . Extrapolating to compare the degree of dilution bias between AA and EA populations , we expect changes in linkage disequilibrium across populations to be reflected by changes in relative effect size:Assuming the effect size of the functional variant is the same in both populations , this reduces to: The above equation allows us to directly compare the observed distribution of relative effect sizes at the tagSNPs in AA and EA ( ) against the relative strength of tagging in AA and EA ( ) . Considering the subset of index tagSNPs in regions that were present on the Metabochip , we observed 51 index tagSNPs that fell into 47 independent loci on the Metabochip . We identified the set of SNPs tagged by each index tagSNP at r2>0 . 8 in an EA population [29] , [30] , yielding a total of 1 , 093 tagged SNPs for the 51 index tagSNPs . For each of these 1 , 144 SNPs , we then calculated . Let this represent the expected distribution of differential LD between AA and EA . Next , we calculated for the subset of 40 of the 51 index tagSNPs that replicated at q = 0 . 05 in EA , truncating at 0 if the signs were opposite between populations . These two distributions ( in all 1 , 144 SNPs versus for the 40 index tagSNPs ) were not significantly different by two-tailed t test . Thus , we cannot reject the hypothesis that the observed dilution bias in AA effect sizes at the index tagSNPs is consistent with the observed distribution of differential LD between the two populations . A single-locus example of the potential for differential LD to contribute to diluted effect sizes is shown in Figure 2 . Considering the 12 SNPs showing differential effect size in AA , regions spanning 11 were present on the Metabochip ( Table S3 ) . Before comparison with EA , we compared the observed effect sizes at the index tagSNPs in the full AA sample and the subsample of AAs genotyped on the Metabochip ( AAmchip ) . Three of the index tagSNPs failed to genotype on the Metabochip , leaving eight index tagSNPs for this direct comparison ( Table S4 ) . No significant allele frequency differences were observed between the AAmchip subset and the full AA population , consistent with AAmchip being a representative subsample . A significantly inconsistent and diluted effect size in AAmchip compared to EA was still observed for five of these eight tagSNPs ( p<0 . 05 , Table S4 ) . The index tagSNPs without a significant difference likely reflect reduced power to detect the differential effect size in the AAmchip subsample , as these three index tagSNPs also had the least significant differential effect when comparing the full PAGE AA subpopulation against EA . The Metabochip genotype data allowed us to evaluate regions spanning each of the 11 variants for the underlying contributions of population-specific alleles , differential tagging , and secondary alleles to differential effect sizes . Detailed discussion of each locus is provided in Text S1 . In summary , the 11 SNPs fell in 10 Metabochip regions , so all SNPs in each of the 10 regions were assessed for association with the reported trait in AAmchip . The threshold level for significance within each region was conservatively adjusted for multiple testing by Bonferroni adjustment for the number of SNPs successfully genotyped on the Metabochip within the region , with minor allele frequency greater than 1% in the AAmchip sample . For example , the Metabochip region spanning CETP contained 84 SNPs , so our significance threshold for that region was p<0 . 05/84 = 1 . 1*10−4 . One locus ( APOE ) could not be dissected confidently as LD data for the index tagSNP were not available in EA , and two loci were underpowered to draw strong conclusions , as evidenced by the failure of any variant in the region to show a significantly inconsistent effect with the index tagSNP effect in EA . Among the remaining seven loci , we observed one clear example of a diluted signal consistent with EA-specific functional alleles , either common or synthetic ( Figure 3a ) , and five loci showed patterns consistent with fine-mapping of the index tagSNP bin ( Figure 3d–f , Figure 4a , 4d ) . One of these fine-mapped the EA association to a variant that was not strongly associated with the index tagSNP in EA ( r2<0 . 5 , Figure 3f ) , potentially consistent with a synthetic allele in EA . We also observed statistically significant secondary functional alleles at three loci ( Figure 4 ) . Thus , although the overall pattern of effect dilution in AA is consistent with expectations on the basis of differential LD patterns between AA and EA populations , putative examples of EA-specific alleles and secondary alleles in AA were also observed . A contribution from synthetic alleles cannot be excluded , and may well account for the EA-specific allele at CILP2 ( Figure 3a ) . However , at half of the 10 loci we observed at least one of the tagged SNPs in EA that showed an effect size in AA consistent with the effect size at the tagSNP in EA . These examples of fine-mapping EA signal suggest that at least half of EA GWAS signals tag a common , functional variant . The observed excess of dilution effects in AA ( as compared to other non-EA populations ) suggests that African-descended populations will be the most useful single subpopulation for fine-mapping of EA GWAS associations , although the significant trend toward excess dilution in HA and NA populations ( Table S5 ) suggests that trans-ethnic fine-mapping may prove more powerful than fine-mapping with any single non-EA population . In conclusion , we have assessed the generalization of GWAS associations from EA populations across five clinically relevant traits , in five non-EA populations . Our results demonstrate that although most EA GWAS findings can be expected to show an effect in the same direction for non-EA populations , a significant fraction of GWAS-identified variants from EA will exhibit differential effect sizes in at least one non-EA population , and these differential results will be far more frequent in the AA population . These findings suggest that expanded GWAS and fine-mapping efforts focused on non-EA populations , especially AA , will substantially enhance our understanding of the genetic architecture of common traits within non-EA populations . It will be particularly important to extend GWAS discovery efforts to non-EA populations if genetic risk prediction models using tagSNP genotypes demonstrate clinical utility , because risk estimates derived from European GWAS clearly generalize imperfectly to non-EA populations . Our analyses suggest that variable LD in its many guises accounts for much of the heterogeneity of effect size at index tagSNPs , rather than any “true” differences in effect size between populations for the functional variants that were tagged . Thus , risk models derived directly from genotypes at functional variants ( rather than tagSNPs ) may generalize more effectively to non-EA populations .
Traits considered were those for which more than 10 GWAS-identified variants were genotyped in the first year of PAGE . Variants considered for this analysis included 13 previously reported to associate with body mass index , 20 for type 2 diabetes , 27 for HDL , 19 for LDL , and 14 for triglycerides . Eleven of these GWAS-identified variants were previously reported to associate with more than one trait in EA ( Table S1 ) , so we constrained the analysis of each such SNP to whichever trait had the most significant association ( smallest p value ) in the PAGE EA population , leaving a panel of 82 unique variants . Because highly correlated SNPs might overweight specific results toward a specific trait or gene , we extracted a subset of minimally correlated index GWAS-identified variants from this panel of 82 . At each step , we added the SNP with the most significant association in PAGE EA to a list of index SNPs , and then filtered the remaining SNPs not yet in the index list to exclude those exceeding r2 = 0 . 2 in the PAGE EA population with any index tagSNP . The panel of 82 SNPs was recursively filtered in this manner , leaving a final panel of 69 index SNPs , each of which was minimally associated with any other index SNP in the PAGE EA ( r2<0 . 2 ) . One additional SNP ( rs11084753 ) was removed from the analysis due to concerns regarding power to replicate , leaving a final panel of 68 index SNPs for analysis , including seven index SNPs for BMI , 18 for HDL , 15 for LDL , nine for triglycerides , and 19 for T2D ( Table S2 ) . Power estimates are taken directly from Fesinmeyer et al . [13] for BMI , Dumitrescu et al . [15] for lipids , and Haiman et al . [14] for T2D . For details , see the original publications . In order to assess the generalization of effects to each population , we used effect sizes ( β ) and standard errors derived from minimally adjusted ( age , sex , and study ) , ancestry-specific meta-analyses described in the primary PAGE publications [15] , [13] , [14] . Using these data , we tested two hypotheses: first , that the GWAS-identified variant has no effect in the non-EA population ( i . e . , the coefficient ßpop = 0 in a linear or logistic regression model ) , and second , that the effect size in the non-EA population is the same as the effect size in EA ( ) . The first hypothesis was tested by assuming the estimate is normally distributed ( which is reasonable as sample sizes exceeded 500 for all populations ) and calculating the probability that given and the standard error of . The second hypothesis was tested by defining and calculating the probability that , again assuming to be normally distributed . These tests are all carried out at a nominal significance level of 0 . 05 , as we see them as the ( single ) test that an investigator may carry out to validate a result first observed in EA in another ethnic group , and then significance was assigned using the Benjamini-Hochberg method at a false discovery rate of 5% . A reasonable concern in these analyses is that population stratification can distort effect size estimates in some circumstances . Some of the effect sizes from trait-specific PAGE manuscripts were not adjusted for genetic ancestry , due to either availability of data [15] or informed consent in specific populations [13] , [14]; where available and allowed we have used the ancestry adjusted effect sizes . Both ancestry adjusted and unadjusted data were available for the PAGE obesity analysis [13] , where ancestry adjustment did not significantly alter effect size estimates .
|
The number of known associations between human diseases and common genetic variants has grown dramatically in the past decade , most being identified in large-scale genetic studies of people of Western European origin . But because the frequencies of genetic variants can differ substantially between continental populations , it's important to assess how well these associations can be extended to populations with different continental ancestry . Are the correlations between genetic variants , disease endpoints , and risk factors consistent enough for genetic risk models to be reliably applied across different ancestries ? Here we describe a systematic analysis of disease outcome and risk-factor–associated variants ( tagSNPs ) identified in European populations , in which we test whether the effect size of a tagSNP is consistent across six populations with significant non-European ancestry . We demonstrate that although nearly all such tagSNPs have effects in the same direction across all ancestries ( i . e . , variants associated with higher risk in Europeans will also be associated with higher risk in other populations ) , roughly a quarter of the variants tested have significantly different magnitude of effect ( usually lower ) in at least one non-European population . We therefore advise caution in the use of tagSNP-based genetic disease risk models in populations that have a different genetic ancestry from the population in which original associations were first made . We then show that this differential strength of association can be attributed to population-dependent variations in the correlation between tagSNPs and the variant that actually determines risk—the so-called functional variant . Risk models based on functional variants are therefore likely to be more robust than tagSNP-based models .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
Generalization and Dilution of Association Results from European GWAS in Populations of Non-European Ancestry: The PAGE Study
|
Orientation preference maps ( OPMs ) are present in carnivores ( such as cats and ferrets ) and primates but are absent in rodents . In this study we investigate the possible link between astrocyte arbors and presence of OPMs . We simulate the development of orientation maps with varying hypercolumn widths using a variant of the Laterally Interconnected Synergetically Self-Organizing Map ( LISSOM ) model , the Gain Control Adaptive Laterally connected ( GCAL ) model , with an additional layer simulating astrocytic activation . The synaptic activity of V1 neurons is given as input to the astrocyte layer . The activity of this astrocyte layer is now used to modulate bidirectional plasticity of lateral excitatory connections in the V1 layer . By simply varying the radius of the astrocytes , the extent of lateral excitatory neuronal connections can be manipulated . An increase in the radius of lateral excitatory connections subsequently increases the size of a single hypercolumn in the OPM . When these lateral excitatory connections become small enough the OPM disappears and a salt-and-pepper organization emerges .
The cortex is the outermost layer of cerebral tissue , composed of neuronal cell bodies and protoplasmic astroytes . The neurons in the cortex are arranged in columns , and the neurons in each column usually respond to similar features . In the macaque these columns , known as microcolumns or minincolumns have a density of 1270 minicolumns per mm2 , with each minicolumn having around 142 pyramidal cell bodies [1] . Now the 3-d volume of cortical tissue could be locally approximated as a 2-d sheet of nodes , with a single node representative of all the neurons within a particular column . With this approximation it becomes possible to describe a 2-d map in the neuronal space with each node responding to a particular feature in the stimulus space . A number of such stimulus modality-specific feature maps are topographic in nature , meaning that features that are similar in the stimulus space are mapped onto neighboring locations in the cortical space . A few examples include the tactile map in the primary somatosensory cortex [2] , the whisker map in the barrel cortex [3] , and the orientation , direction and retinotopic maps in the primary visual cortex [4] . Understanding the mapping function allows prediction of what features a particular neuron will respond to . A model which simulates the development of such maps , would aid in understanding which factors contribute to the development of such features maps . These factors could include internal factors such as the connectivity between the nodes , or the available area of the cortex onto which the features are to be mapped . Similarly features of the stimuli used for training the model themselves act as external factors . Self-organizing maps ( SOMs ) have been used extensively to simulate the development of cortical maps [5–12] . A SOM has two constraints: coverage and continuity . Optimal coverage implies all input stimuli are mapped evenly on to the output space . Continuity implies that neighboring neurons in the output space respond to similar stimuli . The SOM uses local learning rules in order to optimize coverage and continuity . A biologically realistic variant of SOM , namely the Gain Controlled Adaptive Lateral ( GCAL ) , has been used to investigate the factors involved in the development of a number of feature maps in the primary visual cortex ( V1 ) [13] . The GCAL model consists of sheets of neurons . Each neuron in each layer could have 3 kinds of connections , each of which is trained using a normalized Hebbian learning rule: A common feature in most SOMs is the presence of a mechanism by which neighboring neurons respond to similar features whereas those further away respond to dissimilar ones . The GCAL model achieves this by having short range excitatory connections , and longer range inhibitory connections . However in V1 inhibitory connections are local ( short range ) and may dominate responses [14] , whereas the long range connections are excitatory . The effective long range inhibition is achieved by excitatory neurons synapsing onto inhibitory neurons which in turn synapse onto other neurons in its vicinity . For high contrast stimuli , it is known that the long range connections are in effect ( multi-synapse ) inhibitory in nature [15] . From a computational perspective , the radius of the effectively short range excitatory connections is important in determining the size of the orientation hypercolumn [6 , 16] . In the absence of any excitatory connections , with Hebbian trained afferent connections and anti-Hebbian trained lateral connections , a sparse representation yielding independent components of the training set is realized [17] . This implies that an OPM will give way to a salt-and-pepper organization , without a smooth shift in orientation preference among neighboring neurons , in the absence of lateral excitatory connections . OPMs are present in carnivores ( such as cats and ferrets ) and primates but absent in rodents [18] . The term ‘salt-and-pepper’ was originally used to describe the maps seen in rodents , since the orientation preference of neighboring neuronal columns appeared to be uncorrelated and resembled a random pattern . However , recent experimental evidence suggests that the map is pseudo-random and exhibits some local similarities in orientation preference [19] . We hypothesize a possible link between astrocytic arbors and presence of OPMs and try to show that larger astrocytic arbors are more conducive to the generation of OPMs . We investigate the above hypothesis using computational modeling . We propose a GCAL model having 2 V1 layers: one representative of neurons , whereas the other of astrocytes . The synaptic activity of V1 neurons is given as input to an astrocyte layer . The activity of the astrocyte layer is now used to modulate bidirectional plasticity of lateral excitatory connections in the V1 layer . By simply varying the radius of astrocytes , the effective extent of lateral excitatory neuronal connections can be manipulated . An increase in the effective radius of lateral excitatory connections subsequently increases the size of a single hypercolumn in the OPM . When these effective lateral excitatory connections become small enough the OPM disappears and a salt-and-pepper organization emerges . Hubel and Wiesel proposed that the emergence of orientation preference in principal ( layer 4 ) neurons in the primary visual cortex is primarily due to the spatial arrangement of LGN afferent connections [20] , though the effect of recurrent connections is now also clear . This contribution of afferent feed-forward connections is also emphasized by Paik and Ringach , who attribute the development of orientation preference maps across species to the Moire interference patterns created due to the spatial arrangement of Retinal Ganglion Cells ( RGCs ) [21] . While the contribution of feed forward connections to map formation is undeniable , as verified by a number of experiments , the contribution of recurrent lateral connections between cortical columns is also prominent . At the level of columns , rather than at the level of single neuron , it is known that for high contrast inputs , due to the recruitment of local inhibitory inter-neurons , long range lateral connections are predominantly inhibitory in nature [15 , 22 , 23] . This configuration of lateral connections is essential for map formation [7] . What shapes the lateral circuitry in cortical networks ? Are there mechanisms which could ensure that short range connections are excitatory , whereas as the long range connections are in effect ( considering the contribution of interneurons ) predominantly inhibitory ? We hypothesize that protoplasmic astrocytes could play a key role in this regard . Although there are a number of mechanisms by which astrocytes and neurons communicate with each other [24–27] , not all these mechanisms contribute to long term plasticity , crucial for the development of cortical maps . It must however be noted that there are a number of ways in which astrocytes could influence long term plasticity . These mechanisms could be summarized as follows: In each of these mechanisms the effective synaptic strength is influenced by astrocytic activity . NMDA-dependent LTP/LTD is known to be a function of the postsynaptic calcium influx [34] . The postsynaptic calcium influx is likely dependent on astrocytic activity as well . The astrocytic influence could be abstracted using a plasticity or learning rule ( such as a BCM curve ) , where the threshold controlling LTP vs . LTD is dependent on the astrocytic activity . The lateral excitatory connections in the modified GCAL model are modeled in such a manner .
The Gain Control , Adaptation , Laterally Connected ( GCAL ) model , has been used to develop stable and robust orientation maps [35] . This model builds on the LISSOM model and has , as the name suggests , a mechanism which ensures gain control of input activations and homeostatic adaptation of weights . The model has 3 layers: a photo-receptive input layer , an ON/OFF LGN layer and a V1 layer . The activity of the ON/OFF LGN layer is given as L for a node i , j in the layer . L i , j ( t + 1 ) = f ( γ o ∑ a , b x a , b ( t ) C i j , a b k + γ s ∑ a , b L i , j ( t ) C i j , a b s ) ( 1 ) where ( a , b ) denotes a neuron in the receptive field of the ( i , j ) th neuron in the output layer , with input given as xab; Cij , ab represents the weight from the ( a , b ) th neuron to the ( i , j ) th neuron . A constant multiplier to the overall strength is given by γo; γs represents the gain-control . The weights Cij , ab are defined as a difference of Gaussians . C i j , a b = 1 Z c e x p ( - ( a - i ) 2 + ( b - j ) 2 2 σ c 2 ) - 1 Z s e x p ( - ( a - i ) 2 + ( b - j ) 2 2 σ s 2 ) ( 2 ) where Zc , and Zs denote the normalization factors , σc , and σs regulate the width of the gaussians . The term C i j , a b s denotes the lateral inhibition received from other ON/OFF units . C i j , a b s = 1 Z s e x p ( - ( a - i ) 2 + ( b - j ) 2 2 σ c 2 ) ( 3 ) The firing rate of a V1 neuron is dependent on only 3 kinds of inputs , namely: afferent inputs from the LGN ( Lab ( t − 1 ) ) , lateral effectively excitatory inputs , and lateral effectively inhibitory inputs . Thus the firing rate ( yij ( t ) ) is given as: y i j ( t ) = f ( p ∑ a , b A i j , a b L a b ( t - 1 ) + q ∑ k , l E i j , k l y k l ( t - 1 ) - r ∑ k , l I i j , k l y k l ( t - 1 ) ) ( 4 ) where p , q , r are scaling factors; Aij , ab is the afferent weight from neuron ( a , b ) to neuron ( i , j ) ; Eij , kl is the lateral excitatory weight from neuron ( k , l ) to neuron ( i , j ) and similarly Iij , kl is the lateral inhibitory weight from neuron ( k , l ) to neuron ( i , j ) . The function f is a half wave rectifier in order to ensure that the activations are positive with a variable threshold point given as ρ . The activations yij ( t ) are allowed to adapt in a homeostatic fashion . The output activity yij and the threshold ρ are adapted as follows: y ¯ i j ( t ) = ( 1 - β ) y i j ( t ) + β y ¯ i j ( t - 1 ) ( 5 ) ρ ( t ) = ρ ( t - 1 ) + λ y ¯ i j ( t ) - μ ( 6 ) where β is the smoothing parameter and λ is the homeostatic learning rate; y ¯ i j ( t ) is initialized to the average V1 activity ( μ ) . In order to model astrocytic activation we simulate an additional layer whose input is the synaptic activity ( gs ) present at each node of the V1 layer . Thus the activation of a single node in this astrocyte layer is given by Sij . g s i j ( t ) = p ∑ a , b A i j , a b L a b ( t - 1 ) + q ∑ k , l E i j , k l y k l ( t - 1 ) - r ∑ k , l I i j , k l y k l ( t - 1 ) ( 7 ) S i j ( t ) = ∑ i , j ∈ R a s t r o g s i j ( t - 1 ) ( 8 ) where the radius of the astrocyte is given as RAstro . There is some debate regarding the precise nature of GABA induced calcium oscillations in the astrocyte and the subsequent gliotranmission [36] . Hence we run an additional simulation which does not consider the effect of GABA induced gliotransmitters . Now the activation of a single node in the astrocyte layer is given by Sij . g s i j ( t ) = p ∑ a , b A i j , a b L a b ( t - 1 ) + q ∑ k , l E i j , k l y k l ( t - 1 ) ( 9 ) S i j ( t ) = ∑ i , j ∈ R a s t r o g s i j ( t - 1 ) ( 10 ) The lateral inhibitory and afferent weights are trained using the same normalized Hebbian rule given by: w i j , m n ( t + 1 ) = w i j , m n ( t ) + η y i j ( t ) P m n ( t ) ∑ m n ( w i j , m n ( t ) + η y i j ( t ) P m n ( t ) ) ( 11 ) where Pmn is the generalized notation for the pre-synaptic activity originating from the neuron ( m , n ) ; η is the learning rate . These learning rates can be different for each of the connections: ηA , ηE and ηI are the learning rates for the afferent , excitatory and inhibitory connections respectively . However the lateral excitatory connections adapt using a variant of the BCM rule with a threshold function θ being a function of the astrocytic activation at the corresponding node . It has been previously proposed that astrocytes introduce metaplasticity by shifting the BCM curve [31] . E i j , k l ( t + 1 ) = E i j , k l ( t ) + η E y i j ( t ) ( y i j ( t ) - θ i j ) y k l ( t ) ( 12 ) θ i j = ( 1 - S i j ) ( 13 ) Astrocytes communicate with each other via gap junctions; however only distal branches are connected , resulting in astrocytic microdomains with less than 10% overlap [37] . The gap junctions could be modeled using Gaussian random lateral excitatory connections to the 8 nearest neighboring nodes . A schematic of the model is shown in Fig 1 The parameters used for the GCAL model are a superset of those used in the standard LISSOM model . The complete list of parameters are given in Table 1 . The simulations are performed using the Topographica simulator [38] .
We vary the astrocytic radius and observe the changes in the orientation map developed . The experimentally reported astrocytic radii are estimated using the Glial fibrillary acidic protein ( GFAP ) as the astrocytic marker . However , the GFAP marked region accounts for only 15% of the actual astrocytic volume . Hence we scale the astrocytic radii by a factor of 2 in the simulations . The model is trained for 10000 iterations . The training regime consists of elongated 2-dimensional Gaussians with centers and orientations drawn from a uniform random distribution . The astrocytic radius is varied and the corresponding orientation maps developed are studied ( Figs 2 and 3 ) . It is observed that on reducing the astrocytic radius , the periodicity of the map increases and the width of a single hypercolumn decreases . Thus in a given area of cortical tissue 3 x 3 mm , the number of orientation hypercolumns would increase as we reduce the astrocytic radius . The neuronal and astrocyte maps developed have similar orientation preferences which could be quantified by their stability index . The stability index between the astrocytic and orientation maps is shown in Fig 4 . These results demonstrate that the astrocyte radius has a profound effect on OPM formation . The development of a few of these maps and their stability indices across iterations are shown in Figs 5 , 6 , and 7 . The V1 orientation preference map is probed at 250 , 500 , 750 , 1000 , 2500 , 5000 , 7500 and 10000 iterations . It is observed that the map developed becomes stable after a few initial iterations , as quantified by the corresponding stability indices . These results demonstrate the model develops stable orientation maps . We also simulate 2 additional conditions which could effect the development of the orientation maps: ( 1 ) Considering there is no GABA induced gliotransmission: Since the effect of GABA induced calcium oscillations is not well understood in literature , we also simulate the map development ignoring the corresponding term as described in Eq 9 . ( 2 ) Considering the effect of gap junctions in the astrocyte layer: The basic simulation does not consider the effect of gap junctions among astrocytes . As described in the methods section , we introduce gap junction by considering excitatory connections among the nearest neighbors in the astrocyte layer . The maps formed for these 2 conditions are shown in Figs 8 and 9 respectively . The maps developed using all 3 conditions ( basic , no GABA , Gap junctions ) appear visually similar and their features , which are further quantified ( See Fig 10 ) , show a similar trend . These results indicate that the correlation between astrocyte radius and hypercolumn widths is robust for all the conditions considered . The number of pinwheels observed in the neuronal ( V1 ) orientation map in the simulated region ( 3 x 3 mm ) is shown in Fig 10 ( A ) . As expected , the number of pinwheels falls with increasing astrocytic radius . The number of pinwheels per hypercolumn remain approximately constant , centered around π for the maps in which a clear orientation preference map ( OPM ) structure is present ( Fig 10 ( B ) ) . However for smaller astrocytic radii the map begins to disintegrate . These results strengthen the hypothesis that the astrocytic radii influence the formation of orientation maps . For higher astrocytic radii the number of pinwheels per hypercolumn stabilizes to values around π . This result is in keeping with experimental findings which show that the number of pinwheels per hypercolumn is a constant π across species [39] . The trend observed in the simulated widths of the hypercolumn and the corresponding astrocytic radii are comparable with the scant experimental evidence available as shown in Fig 10 ( C ) . The transition from a salt and pepper kind of map to a smooth orientation map could be quantified using 2 methods: ( 1 ) Change in the number of pinwheels/ hypercolumn: Experiment results indicate that the number of pinwheels/ hypercolumn remains a constant across species , even with differing hypercolumn widths [39] . Thus , if such a ratio is no longer maintained , the map developed no longer resembles a smooth orientation preference map . However , the map developed is also not truly random since there might be local patches with similar orientation preference . A recent study has shown that in rodents the map only appears to be random , and has significant local orientation similarity [19] . ( 2 ) Local similarity in orientation preference: This method quantifies the local smoothness of the map developed . The mean angle of separation between the orientation preference of a node and all others within a predefined radius of interest is computed and compared for different astrocytic radii . We then compare the results using the 2 methods and observe that a sharp transition between salt and pepper and a smooth orientation maps is absent ( Fig 11 ) . Rather , an intermediate state which exhibits local patches of orientation similarity , but lacks the features of a true orientation map is seen .
A fascinating feature of orientation mapping is that not all species display a smooth transition in orientation preferences as we probe along the cortical surface . Rodents , in particular have neuronal columns which are orientation specific but arranged in a seemingly randomized fashion across V1 . This kind of organization is referred to as a salt and pepper configuration . The presence or absence of OPMs and their potential consequences for information processing is a topic of current interest . Another interesting fact in those species which do have OPMs is that the size of the hypercolumn varies from species to species . However the number of pinwheels per hypercolumn appears to remain constant across species . Self organizing mechanisms have been extensively utilized to model OPMs . These models rely on a mechanism that ensures that neighbouring neuronal columns respond to similar features , whereas distant ones to different features . This is invariably implemented by invoking local excitatory and larger inhibitory connections . However cortical inhibitory connections are known to be short range , whereas excitatory ones are longer ranged laterally . These long ranged inhibitory connections have been explained away as long ranged excitatory neurons recruiting local inhibitory neurons , such that the net effect is inhibitory . However , a mechanism that ensures that short range connections are effectively more excitatory than inhibitory has proven elusive . The arguments summarized above have been discussed in detail by Swindale [7] . He postulates the possibility of extracellular diffusion of chemical messengers mediating this short range excitatory connectivity . However there are a number of issues with the diffusion hypothesis . Firstly diffusion , a passive process , would ensure roughly similar excitatory radii across species and would thus imply by extension similar widths of orientation hypercolumns across species . In reality the widths of hypercolumns vary widely across species . Rodents do not have a smooth topographical variation in orientation preference and hence do not have defined hypercolumns [18] . Thus diffusion alone would not explain the variation in hypercolumn widths . Secondly the pyramidal apical dendrites , which are used to define the width of a minicolumn ( also called microcolumn ) are roughly the same ( ≈ 30μm ) for the rhesus macaque and the rat [40] . Thus the chemical messenger which diffuses should have similar effects at the level of cortical columns . However as stated earlier this again does not hold true . Astrocytes are known to regulate both excitatory and inhibitory cortical circuits , via a combination of glutamate and GABA re-uptake by transporters , gliotransmitter release , and regulation of neuronal excitability [27–29 , 31 , 41 , 42] . Indeed , optogenetic astrocyte calcium activation modulates the excitatory-inhibitory balance and increases response selectivity of excitatory neurons within local cortical microcircuits [28] . Thus the extent of astrocyte influence may directly influence the range of local influence in the cortex . As mentioned earlier , protoplasmic astrocytes are thought to contribute to metaplasticity [30] . As shown in Fig 12 , astrocytes release gliotransmitters which are known to shift a BCM like curve to the left , implying greater LTP for lower postsynaptic firing rates in excitatory neurons [31] . This constitutes a greater increase in synaptic strength for those synapses in the vicinity of the gliotransmitter releasing astrocyte . Now , astrocytes are understood to release gliotransmitters in correlation with their internal calcium levels [30] . Glutamate in the synaptic cleft , either via receptors or transporters , mediates the calcium levels in the engulfing astrocyte . Astrocytes associated with synapses corresponding to those layer 4 pyramidal neurons , which receive direct thalamic input , would have greater calcium levels as compared to other astrocytes . Hence the metaplasticity induced in the neighboring excitatory synapses in the domain of these astrocytes would also be more pronounced . Over time , this would lead to a greater excitatory drive for those neurons in the vicinity of the neuron receiving the direct thalamic input , as shown in Fig 13 . We hypothesize that astrocytes influence local synapses and specify the radius of lateral excitatory connections . This in turn influences the size of hypercolumns across species . A comparative table specifying hypercolumn widths and astrocyte radii is specified in Table 2 . In the standard GCAL model there is a constraint placed on the maximum radius of the lateral excitatory connections [35] . Indeed , this constraint is necessary to ensure that the lateral excitatory connections are shorter in range than the lateral inhibitory ones . Such a configuration is essential for the development of the self organized orientation map . In our proposed model the lateral excitatory connections have no defined maximum radius . The limit on the lateral excitatory connections is implicitly imposed due to the fact that the astrocytes control the BCM threshold of the excitatory neuronal synapses within their ( astrocyctic ) radius of influence and ensures LTP . For synapses outside the astrocytic radius , the threshold is such that LTD occurs and these connections are pruned off automatically . As a computational principle , any mechanism that can control the lateral excitatory radius with respect to the lateral inhibitory radius , could produce similar maps as those shown in this manuscript . However , several lines of evidence indicate that astrocytes are strongly involved in this mechanism . First , astrocytes have been shown to regulate the excitatory to inhibitory balance in local neuronal circuits [28] . Importantly , astrocytes express transporters for both glutamate and GABA , and can thus regulate the strength of both excitatory and inhibitory synaptic transmission . Second , they have a significant role in regulating local synaptic plasticity , in particular local neuronal excitation , via a range of mechanisms that include modulation of NMDA receptors as well as integrating other plasticity-mediating neuromodulators such as acetylcholine , noradrenaline [43 , 44] . Together , these effects are well placed to implement the BCM rule . Third , astrocytes are known to form microdomains with less than 10% overlap [37] . Thus , the astrocytic organization automatically results in the formation of local domains , which are influenced by these transmitters/modulators . Fourth , the radii of these local domains are known to be less than the effective lateral inhibitory radius , thus resulting in the required short range excitation and longer range inhibition . We simulate the development of orientation maps with varying hypercolumn widths , by simply varying the radius of astrocytic connections using the LISSOM model with an additional layer simulating the astrocytic activation . We observe that increasing the astrocytic radius , and thereby the effective radius of lateral excitatory connections in the V1 neuronal layer , the width of the hypercolumn developed shows a proportionate increase . When the effective lateral excitatory radius is reduced so as to almost prevent any similarity in orientation preference of neighboring neurons , the OPM disappears and a salt and pepper configuration of neuronal arrangement of orientation preference is seen .
|
Columns of neurons in the primary visual cortex ( V1 ) are known to be tuned to visual stimuli containing edges of a particular orientation . The arrangement of these cortical columns varies across species . In many species such as in ferrets , cats , and monkeys a topology preserving map is observed , wherein similarly tuned columns are observed in close proximity to each other , resulting in the formation of Orientation Preference Maps ( OPMs ) . The width of the hypercolumns , the fundamental unit of an OPM , also varies across species . However , such an arrangement is not observed in rodents , wherein a more random arrangement of these cortical columns is reported . We explore the role of astrocytes in the arrangement of these cortical columns using a self-organizing computational model . Invoking evidence that astrocytes could influence bidirectional plasticity among effective lateral excitatory connections in V1 , we introduce a mechanism by which astrocytic inputs can influence map formation in the neuronal network . In the resulting model-generated OPMs the radius of the hypercolumns is found to be correlated with that of astrocytic arbors , a feature that finds support in experimental studies .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"cell",
"physiology",
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"nervous",
"system",
"astrocytes",
"junctional",
"complexes",
"vertebrates",
"electrophysiology",
"neuroscience",
"macroglial",
"cells",
"mammals",
"animals",
"gap",
"junctions",
"synaptic",
"plasticity",
"neurotransmitters",
"neuronal",
"plasticity",
"developmental",
"neuroscience",
"animal",
"cells",
"glial",
"cells",
"gamma-aminobutyric",
"acid",
"biochemistry",
"cellular",
"neuroscience",
"rodents",
"eukaryota",
"cell",
"biology",
"anatomy",
"synapses",
"physiology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"afferent",
"neurons",
"amniotes",
"neurophysiology",
"organisms"
] |
2017
|
The influence of astrocytes on the width of orientation hypercolumns in visual cortex: A computational perspective
|
We report a systematic review and meta-analysis of research using animal models of chemotherapy-induced peripheral neuropathy ( CIPN ) . We systematically searched 5 online databases in September 2012 and updated the search in November 2015 using machine learning and text mining to reduce the screening for inclusion workload and improve accuracy . For each comparison , we calculated a standardised mean difference ( SMD ) effect size , and then combined effects in a random-effects meta-analysis . We assessed the impact of study design factors and reporting of measures to reduce risks of bias . We present power analyses for the most frequently reported behavioural tests; 337 publications were included . Most studies ( 84% ) used male animals only . The most frequently reported outcome measure was evoked limb withdrawal in response to mechanical monofilaments . There was modest reporting of measures to reduce risks of bias . The number of animals required to obtain 80% power with a significance level of 0 . 05 varied substantially across behavioural tests . In this comprehensive summary of the use of animal models of CIPN , we have identified areas in which the value of preclinical CIPN studies might be increased . Using both sexes of animals in the modelling of CIPN , ensuring that outcome measures align with those most relevant in the clinic , and the animal’s pain contextualised ethology will likely improve external validity . Measures to reduce risk of bias should be employed to increase the internal validity of studies . Different outcome measures have different statistical power , and this can refine our approaches in the modelling of CIPN .
Chemotherapy-induced peripheral neuropathy ( CIPN ) is a disabling side effect of many frequently used and effective cancer chemotherapeutic agents and is known to impair daily function and diminish quality of life [1] . Frequently used chemotherapeutic agents reported to cause neurotoxic effects include platinum derivatives , taxanes [2] , vinca alkaloids , epothilones , and also newer agents ( e . g . , thalidomide and bortezomib ) [3] . The predominant sensory phenotype in patients exposed to oxaliplatin or docetaxel is distal symmetrical sensory loss affecting both upper and lower extremities . Symptoms of sensory disturbance reported by patients include paraesthesiae , numbness or tingling , and , less frequently , pain and cold allodynia [4] . CIPN can present clinically in 2 distinct forms: acute and chronic . The acute form is a chemotherapy dose-related , and often dose-limiting , polyneuropathy , which in many cases resolves in patients once the chemotherapy ceases . In some patients , this will persist , with other patients only developing symptoms after treatment has finished . A chronic , often painful , distal sensory neuropathy is still present in 33% of patients 1 year after completion of treatment [5] . No preventive or curative disease modifying treatments exist , and therefore there is a pressing need for more effective treatments [6] . Animal models of CIPN are used to investigate the pathophysiology of CIPN and to test potential therapies [7] . Frequently , chemotherapeutic agents are administered to induce a sensory neuropathy , and behavioural tests are used to assess induced sensory phenomena , such as evoked pain , and locomotor activity . Unfortunately , the conventional paradigm for drug development , in which findings are translated from preclinical animal research to clinical treatments , has been characterised by a lack of success [8 , 9] . Metaresearch from preclinical stroke research suggests that limitations in experimental design , conduct , analysis , and reporting—such as failure to carry out blinded assessment of outcome , randomisation and allocation concealment—may be impeding the development of effective therapies [10–13] . This led to the development of evidence-based guidelines for scientists [14] . These recommendations have been highly successful in transforming the reporting of measures to reduce risk of bias in the preclinical stroke field [15] . We have used a systematic review , in which we systematically identify and appraise all available evidence relevant to a predefined research question to provide a complete and unbiased summary of available evidence . We seek to establish the extent to which limited experimental biases influence the preclinical CIPN literature and to provide evidence to inform tactics to increase the scientific validity of this research . Our aim is to provide a systematic overview of research in the field of in vivo animal modelling of CIPN , with a focus on the reporting of pain-related behavioural outcome measures , to provide useful information for preclinical researchers wishing to improve the design of experiments and refine the in vivo modelling of painful neuropathy .
Our initial systematic search ( September 2012 ) identified 33 , 184 unique publications , of which 6 , 506 were identified as reporting in vivo models of painful neuropathy . This screening stage took 18 person months; 180 of these publications reported models of CIPN ( Fig 1 ) . In the updated search ( November 2015 ) , we identified a further 11 , 880 publications . Using machine learning and text mining , we identified 6 , 108 publications as likely to report models of neuropathic pain , and 928 of these reported models of CIPN . In a random 10% sample of screened publications ( n = 1 , 188 ) , the classifier with the best fit—using stochastic gradient descent—had a screening performance of 97% sensitivity , 67% specificity , and 56% precision . Further details of the different machine-learning approaches applied are available [16] . Of the 928 studies identified to report animal models of CIPN , 157 met our inclusion criteria . From both searches , a total of 337 unique publications are included in this review . The rate of new publications per year is shown in S1 Fig . Metadata from the 337 publications included in this study are available on the Open Science Framework ( OSF; https://doi . org/10 . 17605/OSF . IO/ZJEHY ) . To address concerns that the systematic search is dated we performed a cumulative meta-analysis , in a post hoc analysis , of the effect sizes and tau2 estimates ( an estimate of between-study heterogeneity ) , ordered by year of publication . It appears that the data are mature and stable from around 250 studies onwards ( S2 Fig ) . To investigate sources of heterogeneity we divided the reporting of results by type of study ( i . e . , modelling experiments or intervention experiments ) , and by type of outcome measures reported ( i . e . , pain-related behaviours or other behaviours ) . Therefore , we have 4 datasets: ( i ) Data set 1a—modelling of CIPN and reporting pain-related behavioural outcome measures , ( ii ) Data set 1b—modelling of CIPN and reporting other behavioural outcome measures , ( iii ) Data set 2a –effects of interventions in animal models of CIPN and reporting pain-related behavioural outcome measures , and ( iv ) Data set 2b—effects of interventions in animal models of CIPN and reporting other behavioural outcome measures . Across the 337 publications included , we extracted all behavioural outcome measure data . Pain-related outcome measures included evoked limb withdrawal to stimuli ( mechanical , heat , cold , and/or dynamic mechanical touch ) , evoked limb withdrawal and/or vocalisation to pressure stimuli , evoked tail withdrawal to stimuli ( cold , heat , and/or pressure ) , and complex behaviours , e . g . , burrowing activity . Other outcome measures included assessment of locomotor function , memory , reward , and attention . Pain-related and other outcome measures for both modelling and intervention experiments were analysed separately ( Fig 2 ) . The full list of behavioural outcome measures and behavioural tests is given in Tables 1 and 2 . The most frequently reported pain-related outcome measure was evoked limb withdrawal to mechanical stimuli , most frequently assessed using monofilaments ( Table 1 ) . The most frequently reported other behavioural outcome measure was locomotor function , with the rotarod apparatus used in most cases ( Table 2 ) . A total of 306 different interventions were tested ( S3 Fig ) . Most ( 80% ) were only tested in 1 publication , and the most frequently reported interventions were gabapentin , morphine , and pregabalin , which were reported in 26 , 22 , and 11 publications , respectively . The reporting of measures to reduce risk of bias was ‘moderate’ across included studies ( n = 337 ) : 51 . 3% ( n = 173 ) reported blinded assessment of outcome , 28 . 5% ( n = 96 ) reported randomisation to group , 17 . 8% ( n = 60 ) reported animal exclusions , 2 . 1% ( n = 7 ) reported the use of a sample size calculation , and 1 . 5% ( n = 5 ) reported allocation concealment . Across all included studies , 49 . 6% ( n = 167 ) reported a conflict of interest statement , and 96 . 7% ( n = 326 ) reported compliance with animal welfare regulations ( Table 3 ) . The methods used to implement randomisation and blinding , and the methods and assumptions for sample size calculations , were rarely reported: 6 publications reported that animals were randomly allocated to experimental groups using randomly generated number sequences , and 2 publications reported that this was done by block randomisation ( 8 . 3% of those that reported randomisation; 8 out of 96 ) . One publication reported that randomisation was performed by ‘picking animals randomly from a cage’ , which we do not consider a valid method of randomisation [17] . Nine publications reported that blinded outcome assessment was achieved by using a different experimenter to perform assessments , and 2 publications reported that a group code was used ( 6 . 4% of those that reported blinding; 11 out of 173 ) . One study reported that allocation concealment was achieved using a coded system ( 20% of those that reported allocation concealment; 1 out of 5 ) . Methods of sample size calculation were reported by 5 publications ( 71 . 4% of those that reported a sample size calculation; 5 out of 7 ) : 3 used published or previous results from the group , and 2 had performed a pilot study to inform sample size calculations . In addition , as a secondary outcome , we abstracted data from modelling experiments using other behavioural outcomes ( locomotor function , memory , reward behaviours , and attention ) . Administration of chemotherapeutic agents led to increased pain-related behaviours compared to sham controls ( −0 . 75 [95% CI −1 . 04 to −0 . 47] , n = 63 comparisons ) . Species did not account for a significant proportion of the heterogeneity ( Q = 3 . 29 , df = 1 , p = 0 . 070 ) , and therefore rats and mice were analysed together . In intervention studies using other behavioural outcomes , administration of interventions led to improvement in other behaviours compared to controls ( 0 . 69 SD [95% CI 0 . 37–1 . 0] , n = 37 comparisons ) . Species did not account for a significant proportion of the heterogeneity ( Q = 0 . 75 , df = 1 , p = 0 . 39 ) . The reporting of details of animal husbandry was low across all included studies ( S1 Table ) . No study reported whether different species were housed in the same room .
There was moderate reporting of measures to reduce risk of bias . Our subgroup analyses did not consistently identify that the reporting of these measures had an impact on experimental findings . It may be that there was insufficient power to test for these associations because of the small number of studies reporting these factors or that there is indeed no association . We also are only able to test the reporting of these measures to reduce risk of bias , and these may differ according to the actual use of these measures in the design , conduct , and analysis of a study . The details of methods used to implement randomisation and blinding and the methods and assumptions for sample size calculation were rarely reported . Despite the inconsistency of our findings , there is substantial empirical evidence of numerous research domains that these details are important to understand the validity of the procedures used [17] , noting that one of the included studies reported that randomisation was achieved by selecting animals at random from the cage . If methods and assumptions were reported , this would allow assessment of the quality of these procedures that report using tools such as those used in clinical systematic reviews [26] and allow for more robust assessments of their impact on research findings . Statistical modelling and meta-analysis have demonstrated that the exclusion of animals can distort true effects; even random loss of samples decreased statistical power , but if the exclusion is not random , this can dramatically increase the probability of false positive results [27] . It has been shown in other research fields that treatment efficacy is lower in studies that report measures to reduce risk of bias [13 , 28–30] . Our assessment of publication bias finds evidence to suggest that global effect sizes are substantially overstated in all data sets except the smallest ( and this is likely due to reduced power to detect publication bias with only 37 studies ) . We observed relative overstatements in effect sizes that ranged from 28% to 56% . Publication bias is a prevailing problem in preclinical research , in which neutral or negative studies are less likely to be published than positive studies [31] . One potential reason for this is the high competition for academic promotion and funding , and few incentives to publish findings from studies in which the null hypothesis was not disproved . Initiatives such as Registered Reports provide one mechanism to support the publication of well-designed , thoroughly executed , and well-reported studies asking important questions regardless of the results . Experimental design of in vivo CIPN studies could be optimised by adopting measures to reduce risk of bias , such as using sample size calculations to ensure that experiments are appropriately powered . It is also important to use a model that best represents the clinical population of interest , for example , using both female and male animals . To help further address the issue of publication bias , we suggest that researchers make available prespecified protocols for confirmatory preclinical studies and publish all results . Others have shown that external validity may be increased by using multicentre studies to create more heterogeneous study samples , for example , by introducing variations in the animal genetics and environmental conditions ( housing and husbandry ) between laboratories , an approach that may be useful in pain modelling . One approach that would help optimise experimental design is to use the Experimental Design Assistant ( EDA; https://eda . nc3rs . org . uk/ ) , a free resource developed by the NC3Rs , whereby researchers create a record of their experimental design [32] . The output from the EDA could then be uploaded to the OSF as a record for transparency . There are opportunities to reduce waste and maximise the information gained from in vivo models of pain studies . This would require open and transparent reporting of results . For example , for complex behaviours , the online dissemination of individual animal video files [33] would allow reanalysis for further behaviours not reported in the original publication . It is interesting to note that although the open field was used in studies included in this systematic review , none of the included studies reported thigmotaxis , an outcome measure reported in other preclinical pain research . Sharing open field video files would allow this outcome to be assessed from previously conducted experiments . Our exemplar power calculations of the most frequently reported behavioural outcome measures highlight the substantial variability in the statistical performance of different outcome measures . Using these results , it is possible to rank the different pain-related behavioural tests according to how many animals are required per group as effect size or SD increases or decreases . This allows researchers to evaluate the sensitivity of their estimates of numbers required compared with variations in the effect sizes or variance achieved . Along with other factors , such as clinical relevance , these results can inform the choice of outcome measure in study design by allowing researchers to select outcome measures that require fewer animals . The results of our systematic review show increasing rates of publications of experiments using animal models of CIPN . Between the initial search in 2012 and the updated search in 2015 , the number of relevant publications increased by 89% . The high publication accrual rate is not unique to this field but is the case across clinical [34] and preclinical research; this makes it challenging for researchers and consumers of research to keep up to date with the literature in their field . This systematic review of preclinical models of pain uses machine learning and text mining and demonstrates the usefulness of these automation tools in this field . Conducting a systematic review is time and resource intensive , and the rate of publication of new primary research means that systematic reviews rapidly become outdated . This review is limited because the most recent information included was identified in November 2015 . We plan that the present systematic review form a ‘baseline’ systematic review , which can be updated and developed into a living systematic review , i . e . , one that is continually updated as new evidence is published [35] . An important secondary output of this review is the advances made in the use of machine learning to facilitate the automation of systematic reviews of preclinical studies . As new online platforms and tools for machine learning and automation become available , preclinical living systematic reviews become more feasible [36] . Guidelines for living systematic reviews [36] and the use of automation tools [37] have recently been published , and Cochrane has also launched pilot living systematic reviews [38 , 39] . The machine-learning algorithm based on our initial screening had a high sensitivity ( 97% ) and medium specificity ( 67% ) . High sensitivity has a low risk of missing relevant literature . An algorithm with lower specificity is more likely to falsely identify studies for inclusion ( i . e . , false positives ) . As a result , during data abstraction , the 2 independent human screeners excluded many studies identified by the machine for inclusion . We believe that this balance between sensitivity and specificity was appropriate because this reduced the risk of missing relevant studies . A further possible limitation of our study is that we chose to extract behavioural data at the time point at which there was the largest difference between model and sham control animals or treatment and control animals . This time point was chosen to capture information on intervention effects regardless of their half-life . This limits what we can infer regarding the mismatch between timings , but we did also capture information on the first administration of intervention ( relative to induction of the model ) and the last administration . Future studies may use area under the curve approaches to capture response to model induction or drug intervention , but this was not possible for this large data set . There are tools under development for automation of data extraction , which may assist progress in this area [40] . In our meta-analysis , we grouped together the behavioural outcome measures that measure the same underlying biology . For example , in the case of experiments that reported using the grip test , 5 studies reported that the test was used to measure grip strength , and 1 reported that the test was used to measure muscle hypersensitivity [41] . For this reason , in our analysis , we grouped all grip test outcome measures together as a non–pain-related behavioural outcome measures . It is possible that the same tests or similar tests could be used and the same measurements reported as different outcomes; one test may also measure multiple facets of underlying biology . This is one of the challenges when analysing published data , and principle components analyses of large data sets such as these may help identify latent domains of behavioural outcome . We only included studies in which the intervention drug was administered after or at the same time as the chemotherapeutic agent . Future literature reviews may consider drug interventions given before chemotherapeutic agents to determine whether prophylaxis can effectively prevent CIPN . Unfortunately , the reporting of measures to reduce risk of bias was moderate in the studies included in this systematic review , which limits what we can infer from the results . We hope this review will highlight this issue in in vivo modelling of CIPN . Systematic review of animal experiments in other research areas has revealed low reporting of these measures and the negative impact of failure to report these measures across in vivo domains as diverse as modelling of stroke , intracerebral haemorrhage , Parkinson’s disease , multiple sclerosis , and bone cancer pain [15 , 29 , 30 , 42–44] . This has driven change , influencing the development of reporting guidelines [45] , pain modelling specific guidelines [46] , and the editorial policy of Nature Publishing Group [47] . However , requesting that submitting authors complete a reporting guideline without any other intervention is not associated with improved reporting [48] . After an initial review on the efficacy of interleukin-1 receptor antagonist in animal models of stroke highlighted low reporting of measures to reduce risk of bias [49] , a subsequent review identified increased reporting of these measures [15] , increasing the validity and reliability of these results . We hope that there will be a similar improvement in studies reporting the use of animal models of CIPN . We propose that if more studies implement and report measures to reduce the risk of bias , it will be possible to use a GRADE-type analysis to rate the certainty of the evidence of animal studies [50] . At present , any such approach is likely to lead to the majority of evidence being downgraded to the extent that no firm conclusions can be drawn . The measures to reduce risk of bias that we have assessed are largely derived from what is known to be important in clinical trials , and the extent to which these measures are important in animal studies has yet to be fully elucidated . However , reporting of these measures allows users of research to make informed judgments about the fidelity of the findings presented . Equally , it may be that there are other measures that are important in animal studies that we have not considered . A recent study from our group has suggested that using SMD estimates of effect sizes with stratified meta-analysis has a moderate statistical power to detect the effect of a variable of interest when there are 200 included studies but that the false positive rate is low . This means that although we may not have sufficient power to detect an effect , we can have confidence that any significant results observed are likely to be true [51] . This systematic review and meta-analysis provides a comprehensive summary of the in vivo modelling of CIPN . The data herein can be used to inform robust experimental design of future studies . We have identified some areas in which the internal and external validity of preclinical CIPN studies may be increased; using both sexes of animals in the modelling of CIPN and ensuring outcome measures align with those most relevant in the clinic will likely improve external validity . Measures to reduce risk of bias should be employed to increase the internal validity of studies . Power analysis calculations illustrate the variation in group size under different conditions and between different behavioural tests and can be used to inform outcome measure choice in study design .
In September 2012 , we systematically searched 5 online databases ( PubMed , Web of Science , Biosis Citation Index , Biosis Previews , and Embase ) with no language restrictions to identify publications reporting in vivo modelling of CIPN that reported a pain-related behavioural outcome measure . The search terms used for each database are detailed in S1 File . Search results were limited to animal studies using search filters [52 , 53] . Because we anticipated a high accrual rate of new publications , we ran an updated search in November 2015 and used machine learning and text mining to reduce the screening for inclusion workload . This updated search included 4 online databases ( PubMed , Web of Science , Biosis Citation Index , and Embase ) and used an updated animal filter [54] . Biosis Previews was no longer available . We used machine learning to facilitate the screening of publications reporting animal models of CIPN and improve accuracy of the screening process [55] . The screening stage of a systematic review involves ‘including’ or ‘excluding’ publications identified in the search based on their title and abstract , and this was performed by 2 independent reviewers . The publications from our initial search ( with ‘include’/‘exclude’ decisions based on initial dual screening and differences reconciled by a third reviewer; inter-reviewer agreement Kappa = 0 . 95 , standard error [SE] = 0 . 002 ) were used as a training set for machine learning approaches applied to the updated search . Five machine learning groups participated , and 13 classifiers were created and applied to the updated search ( validation set ) [16] . We manually screened 10% of the updated publications ( n = 1 , 188 ) and used this to assess the performance of these classifiers using measures of sensitivity , specificity , and precision as described by O’Mara-Eves and colleagues ( 2015 ) [56] . The reconciled decision of the human reviewers was considered the gold standard . We chose cut-off points such that the sensitivity of each classifier reached 0 . 95 and measured the resulting specificity and precision to choose the classifier that performed best for our data set . To test the performance of the classifiers in the validation set , we used a random number generator to select a 10% random sample , and 2 independent investigators checked these for inclusion or exclusion . From the included studies in the updated search , we used text mining to identify studies reporting animal models of CIPN by searching for specific chemotherapy terms within the title and abstract of the identified publications; the inclusion of these studies was then verified by 2 independent reviewers . We included controlled studies using pain-related behavioural outcome measures that either characterised models of neuropathy induced by chemotherapeutic agents or tested the effect of a drug intervention in such models ( Fig 2 ) . We required that studies report the number of animals per group , the mean , and a measure of variance ( either the standard error of the mean [SEM] or the SD ) . We excluded studies that administered the drug intervention before model induction , administered co-treatments , used transgenic models , or used in vitro models . We assessed the risk of bias of included studies by recording the reporting of 5 measures to reduce risk of bias at the study level: blinded assessment of outcome , random allocation to group , allocation concealment , reporting of animal exclusions , and a sample size calculation [57] . We also assessed the reporting of a statement of potential conflicts of interest and of compliance with animal welfare regulations [57 , 58] . Data were abstracted to the CAMARADES Data Manager ( Microsoft Access , Redmond , WA ) . For all included studies , we included details of publication ( Table 5 ) , animal husbandry , model , intervention , and other experiment details ( Table 6 ) . Outcome data presented graphically were abstracted using digital ruler software ( Universal Desktop Ruler , AVPSoft . com or Adobe ruler ) to determine values . When multiple time points were presented , we abstracted the time point that showed the greatest difference between model and control groups , or the greatest difference between treatment and control groups . If the type of variance ( e . g . , SEM or SD ) was not reported , we characterised the variance as SEM because this is a more conservative approach in meta-analysis , in which studies are weighted in part by the inverse of the observed variance . All data were abstracted by 2 independent reviewers . Publication and outcome level data abstracted by 2 independent reviewers were compared , and any discrepancies were reconciled . For outcome data , SMD effect sizes of individual comparisons were calculated for each reviewer’s extracted data , and when these differed by ≥10% , they were identified for reconciliation . When individual comparisons differed by <10% , we took a mean of the 2 effect sizes and of the variance measure . We separated the data according to those reporting the modelling of CIPN only and those testing the effect of an intervention in a model of CIPN . We analysed all the behavioural outcome measures reported . Behavioural outcome measures were separately considered as ‘pain-related’ or ‘other ( non–pain related ) ’ behavioural outcome measures ( Fig 2 ) . This resulted in 4 data sets: ( 1 ) animal studies modelling CIPN: pain-related behavioural outcome measures ( data set 1a ) , ( 2 ) animal studies modelling CIPN: other behavioural outcomes ( data set 1b ) , ( 3 ) drug interventions in animal models of CIPN: pain-related behavioural outcome measures ( data set 2a ) , and ( 4 ) drug interventions in animal models of CIPN: other behavioural outcomes ( data set 2b ) . Data from individual experiments were extracted from each publication , and these are reported as ‘individual comparisons’ . For each individual comparison , we calculated an SMD effect size . When more than one relevant behaviour was reported in the same cohort of animals , these individual comparisons were aggregated ( ‘nested comparisons’; Fig 2 ) by behavioural subtype , determined by the site of stimulus application ( e . g . , limb or tail ) and the modality of the stimulus used ( e . g . , mechanical or heat ) . Fixed-effects meta-analysis was used to give a summary estimate of these effects in each cohort . Cohort-level effect sizes were then pooled using a random-effects meta-analysis with restricted maximum-likelihood estimation of heterogeneity , in which heterogeneity refers to the variation in study outcomes between studies . When a single control group served multiple comparator ( model or treatment ) groups , their contribution was adjusted by dividing the number of animals in the control group by the number of comparator groups served . The Hartung and Knapp method was used to adjust test statistics and confidence intervals; this calculates the confidence intervals using the following formula: effect size + t ( 0 . 975 , k − 1 ) × SE . Results are presented in SMD units along with the 95% confidence intervals . To provide empirical evidence to inform experimental design and refine modelling of CIPN , we assessed the extent to which predefined study design and study risk of bias characteristics explained observed heterogeneity . We used stratified meta-analysis for categorical variables and metaregression for continuous variables . The purpose of these subgroup analyses is to observe whether studies grouped together describing a similar characteristic ( e . g . , all studies using male animals versus all studies using female animals ) differ in their overall estimates of effects . Such analyses provide empirical evidence of the impact of study design choices and are useful to design future experiments . The study design factors assessed using stratified meta-analysis were animal sex and species , therapeutic intervention , therapeutic intervention dose , methods to induce the model including the chemotherapeutic agent , and type of outcome measure . Because drug dose and route of administration are largely important in the context of the intervention being used , we did not assess the impact of dose or route of administration across different chemotherapeutic agents or drug interventions . We specified a priori that if species accounted for a significant proportion of heterogeneity , we would analyse the effect of study design factors on each species separately . If not , then all data would be analysed together . We also assessed the impact of reporting of measures to reduce bias . We used metaregression to assess the impact of time to assessment ( defined as the interval between first administration of chemotherapeutic agent and outcome measurement ) and time to intervention administration ( defined as the interval between first administration of chemotherapeutic agent and administration of intervention ) . We used a meta-analysis online platform ( code available here: https://github . com/qianyingw/meta-analysis-app ) to perform all meta-analyses . We applied a Bonferroni-Holmes correction for multiple testing that resulted in critical thresholds for significance as follows: in modelling experiments , p < 0 . 01 for study design features and p < 0 . 007 for reporting of measures to reduce risk of bias and measures of reporting; in intervention experiments , p < 0 . 007 for study design features , and p < 0 . 007 for reporting of measures to reduce risk of bias and measures of reporting . To guide sample size estimation for future studies , we performed power calculations for the 6 most frequently reported behavioural tests . To do this , we separately ranked the observed SMD effect size and the pooled SD and , for each , identified the 20th , 50th , and 80th percentile . We then used these values to calculate the number of animals required in 9 hypothetical treatment and control groups . Calculations were based on the two-sample two-sided t test , with 80% power and an alpha value of 0 . 05 . We assessed for potential publication bias by assessing the asymmetry of funnel plots using visual inspection and Egger’s regression [59] . We assessed for the impact of publication bias using Duval and Tweedie’s trim and fill analysis [60 , 61] . We performed these assessments in 4 data sets separately and used individual comparisons rather than summary estimates for each cohort ( Fig 1 ) . In a clinical systematic review of neuropathic pain [18] , selected analgesic agents had been ranked according to their efficacy , as measured by Number Needed to Treat ( NNT ) for 50% pain relief . If preclinical studies included in this review reported use of these agents or their analogues , we ranked the interventions according to their SMD effect size for attenuation of pain-related behaviour . We then assessed the correlation between clinical and preclinical rank using Spearman’s rank correlation coefficient .
|
Many frequently used and effective cancer chemotherapies can cause a disabling side effect that features pain , numbness , tingling , and sensitivity to cold and heat in the extremities known as chemotherapy-induced peripheral neuropathy ( CIPN ) . There are currently no effective therapies to treat or prevent this condition , and animal models have been developed to address this . It is important that experiments using animal models of CIPN are robust and valid if they are to effectively help patients . We used a systematic approach to identify all 337 studies that have been published describing the use of animal models of CIPN . We were able to identify that many studies are imperfect in their experimental design , use only male animals , and assess outcomes with limited relevance to the human condition . Based on a meta-analysis , we provide guidance to the CIPN animal modelling community to guide future experiments that may improve their utility and validity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"statistics",
"metaanalysis",
"drugs",
"experimental",
"design",
"social",
"sciences",
"biological",
"locomotion",
"chemotherapeutic",
"agents",
"animal",
"models",
"oncology",
"research",
"design",
"animal",
"behavior",
"mathematics",
"experimental",
"organism",
"systems",
"oncology",
"agents",
"pharmacology",
"zoology",
"research",
"and",
"analysis",
"methods",
"meta-research",
"article",
"animal",
"studies",
"behavior",
"mathematical",
"and",
"statistical",
"techniques",
"research",
"assessment",
"publication",
"ethics",
"research",
"integrity",
"psychology",
"science",
"policy",
"systematic",
"reviews",
"physiology",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"statistical",
"methods"
] |
2019
|
Animal models of chemotherapy-induced peripheral neuropathy: A machine-assisted systematic review and meta-analysis
|
Genes vary greatly in their long-term phylogenetic stability and there exists no general explanation for these differences . The cytochrome P450 ( CYP450 ) gene superfamily is well suited to investigating this problem because it is large and well studied , and it includes both stable and unstable genes . CYP450 genes encode oxidase enzymes that function in metabolism of endogenous small molecules and in detoxification of xenobiotic compounds . Both types of enzymes have been intensively studied . My analysis of ten nearly complete vertebrate genomes indicates that each genome contains 50–80 CYP450 genes , which are about evenly divided between phylogenetically stable and unstable genes . The stable genes are characterized by few or no gene duplications or losses in species ranging from bony fish to mammals , whereas unstable genes are characterized by frequent gene duplications and losses ( birth–death evolution ) even among closely related species . All of the CYP450 genes that encode enzymes with known endogenous substrates are phylogenetically stable . In contrast , most of the unstable genes encode enzymes that function as xenobiotic detoxifiers . Nearly all unstable CYP450 genes in the mouse and human genomes reside in a few dense gene clusters , forming unstable gene islands that arose by recurrent local gene duplication . Evidence for positive selection in amino acid sequence is restricted to these unstable CYP450 genes , and sites of selection are associated with substrate-binding regions in the protein structure . These results can be explained by a general model in which phylogenetically stable genes have core functions in development and physiology , whereas unstable genes have accessory functions associated with unstable environmental interactions such as toxin and pathogen exposure . Unstable gene islands in vertebrates share some functional properties with bacterial genomic islands , though they arise by local gene duplication rather than horizontal gene transfer .
Genes in animal and plant gene families vary greatly in their phylogenetic stability . Some genes persist as a single copy over a wide phylogenetic range of species , with few or no gene duplications or losses on different evolving lineages . Other genes undergo frequent duplication and loss in a process called birth-death evolution [1 , 2] . Gene duplication and subsequent divergence of the duplicate copies underlie the formation of gene families and are thought to be an important source of genetic complexity and evolutionary change [3–5] . Though many specific examples of stable and unstable genes have been documented , there exists no general explanation for their different patterns of evolution . To explore the evolutionary basis for these patterns , I sought a gene family with characteristics that make it amenable to detailed analysis and interpretation . The vertebrate cytochrome P450 ( CYP450 ) gene superfamily is well suited for the following reasons . First , it is a large family with both phylogenetically stable members and unstable members . Second , the biochemical and organismal functions of most genes in the family have been extensively studied in humans and rodents . Third , their protein products are large and have a well-conserved domain content , providing abundant information for studies of molecular evolution . Finally , strong interest in the family has resulted in relatively high quality annotations of gene structure in a wide range of vertebrate species . The human genome contains approximately 60 CYP450 genes , which encode membrane-bound oxidase enzymes that act on a wide variety of substrates ( recent reviews include [6–8] ) . The CYP450 enzymes can be divided into those that act on endogenous small molecules and those that act on xenobiotic compounds . The endogenous-substrate enzymes function in biosynthesis or catabolism of steroids , sterols , retinoids , prostaglandins , and fatty acids . The xenobiotic-substrate enzymes are expressed primarily in the liver and epithelial tissues and function to defend against environmental toxins and carcinogens . These xenobiotic CYP450 genes have been studied extensively in humans because they comprise one of the major ( and most polymorphic ) activities that determine drug half-life and are thus important in pharmaceutical development [6 , 7] . Xenobiotic CYP450 enzymes have also been studied extensively because in some cases they convert exogenous compounds into active carcinogens , presumably as a negative side-effect of broad substrate specificity [9] . Approximately 22 of the human CYP450 enzymes are thought to act primarily on endogenous substrates and approximately 15 other CYP450 enzymes are thought to act primarily on xenobiotic substrates ( Figure S1 ) . A few of the remaining enzymes are reported to have activity toward both endogenous and xenobiotic substrates; it is unclear whether one or the other activity ( or both ) is their primary function . Finally , there are about ten CYP450 enzymes with no known activities , most of which have been identified only recently from human genome sequence . To study the molecular evolution of the CYP450 superfamily , I considered for analysis about 20 vertebrate genomes in various stages of draft sequencing , assembly , and annotation . Of these , sequence coverage , assembly , and annotation of 11 genomes were sufficiently complete for my needs . Of these 11 , Pan troglodytes was discarded because of its extreme similarity to human , leaving ten nearly complete genomes: six placental mammals , one bird , one amphibian , and two teleost fish . Analysis of gene stability among these genomes reveals that the rate of birth-death evolution among genes acting on xenobiotic substrates is much higher than among those acting on endogenous substrates .
As starting material , I gathered complete sets of predicted CYP450 genes from the genomes of human ( Homo sapiens ) , rhesus macaque ( Macaca mulatta ) , dog ( Canis familiaris ) , cow ( Bos taurus ) , mouse ( Mus musculus ) , rat ( Rattus norvegicus ) , chicken ( Gallus gallus ) , clawed frog ( Xenopus tropicalis ) , zebrafish ( Danio rerio ) , and pufferfish ( Takifugu rubripes ) . All known and predicted splice forms were screened to identify a single canonical coding sequence from each gene for molecular evolutionary analysis ( see Materials and Methods ) . Figure S1 shows a maximum-likelihood protein tree derived from all 628 of these sequences , marked to indicate known or probable functions based on the human proteins . A representative segment of the tree is shown in Figure 1 . In the tree , CYP450 enzymes that act on endogenous substrates are generally represented by a single gene in each organism , and the phylogeny of that gene approximates the known species phylogeny . In some cases there are two related proteins in one or both fish species , presumably resulting from the known whole-genome duplication on the fish lineage [10 , 11] . In all but one case , each orthologous group of genes is strongly separated on the tree from all other CYP450 genes , indicating that they became functionally specialized well before the divergence of teleost fish from mammals ( the single exception is described below ) . In sharp contrast , nearly all CYP450 enzymes that act on xenobiotic substrates are encoded by genes that have undergone active duplication and loss throughout vertebrate evolution , reflected in multiple species- and lineage-specific gene expansions . The tree segment in Figure 1 includes two endogenous-substrate enzymes and one class of xenobiotic-substrate enzymes; they are typical of the complete tree ( Figure S1 ) . Table 1 summarizes gene numbers and variance across species for three groups of CYP450 genes . Among the endogenous-substrate ortholog groups , an expected protein from a specific species was occasionally absent , usually from cow , chicken , or frog . One example is found in Figure 1 ( a CYP46A1 cholesterol 24-hydroxylase ortholog is missing from cow ) . Most of these apparent gene losses probably result from incomplete sequence assemblies or gene annotations . Such apparent gene losses were found most often in the cow , chicken , and frog genomes presumably as a result of relatively immature assemblies and annotations of these genomes . In several cases , translated BLAST nucleotide ( TBLASTN ) searches of the cognate genome clearly indicated the likely presence of part or all of an unpredicted gene ortholog ( see Materials and Methods; Tables S1 and S2 ) . There were also a few cases of apparent species-specific gene duplications outside of fish . For example , there were two predicted rat genes encoding CYP51A1-like proteins ( Figure S1 ) . Although these cases may include undiagnosed pseudogenes , it is reasonable to suppose that occasional gene duplications may be tolerated in these groups . Finally , there was one case in which enzymes that act on endogenous substrates ( corticosteroid precursors ) appear to have become specialized recently and specifically in terrestrial vertebrates ( CYP11B1 steroid 11-β-hydroxylase and CYP11B2 aldosterone synthase ) , probably in connection with resistance to dehydration [12] . Nevertheless , the general picture among endogenous-substrate CYP450 genes is one of striking phylogenetic stability , with very low rates of gene duplication and loss across jawed vertebrates . Among CYP450 enzymes with xenobiotic substrates the picture was dramatically different . Almost without exception , one or more species or lineages had multiple genes that grouped in the tree with one or no genes from other species . The sample shown in Figure 1 is typical of the complete tree . These differences in gene content are far too extensive to be accounted for by incomplete genome sequence and annotation . In particular , the human and mouse genomes are especially mature in sequence and annotation , and the pattern of gene duplication and loss is robustly apparent in comparing these two genomes ( e . g . , Figure 1; Table 2 ) . Extensive bootstrap testing on several subtrees supported the occurrence of independent gene duplications and deletions on both the mouse and human lineages ( Figure S2 ) . In principle , genetic exchange among closely related genes within a species ( concerted evolution ) could explain some aspects of these patterns by causing lineage-specific clades of genes from sequence homogenization rather than recent duplication . In support of this possibility , a recent paper describes evidence that concerted evolution played a role in the divergence of the CYP1A1 and CYP1A2 genes in mammals from their bird orthologs [13] . However there are several reasons to think that this case is unusual . First , all of the syntenic CYP450 gene clusters between human and mouse contained different numbers of genes , indicating that gene duplication or loss must occur ( two examples are shown in Figures S3 and S4 ) . Second , except for a few of the most closely-related gene pairs , nucleotide divergence among duplicate genes was substantial , nucleotide changes were distributed widely across the coding sequence , and regions of higher nucleotide similarity usually coincided with regions of stronger purifying selection in the family in general ( unpublished data ) . Finally , in many cases the tree relationships were inconsistent with concerted evolution being a dominant influence on sequence relatedness ( for example , Figure S5 ) . These results suggest that concerted evolution plays , at most , a minor role in shaping evolution in the family as a whole . Most human and mouse genes for CYP450 enzymes are scattered widely across their genomes , but there are some exceptions . Clusters of genes are found at human chromosome bands 1p33 , 7q22 , 10p13 , 10q26 , and 19q13 . In each of these five human clusters , the CYP450 genes are adjacent , with no other confirmed genes interspersed among them . Within each cluster , all the genes encode closely related proteins with many cases of apparent gene duplication since the split from the rodent lineage . Strikingly , nearly all of the CYP450 enzymes with known xenobiotic substrates were found in these gene clusters ( e . g . , see Figure 1 chromosome positions ) . The remaining genes in these clusters were also phylogenetically unstable but are not yet implicated in xenobiotic detoxification . Each of the human clusters had a syntenic counterpart cluster in mouse ( Figures S1 , S3 , and S4 ) . Very few of the genes in these clusters could be assigned as one-to-one orthologs because of repeated duplication and deletion events on both lineages . These syntenic gene clusters must have originated from a shared ancestral gene or genes , with repeated local gene duplications and losses giving rise to lineage-specific groups of related genes . The patterns of divergence are consistent with single gene duplications that occurred at various times , though multigene duplications followed by loss of most of the duplicate genes cannot be ruled out . Genomic clustering of related genes was also apparent in the other eight genomes , but assembly quality was more problematic , and I did not investigate the patterns in detail . These results are consistent with increased rates of birth-death evolution specific to xenobiotic-substrate genes , with local gene duplications giving rise to clusters of related genes . If exposure to xenobiotic compounds changed over evolutionary time , the CYP450 detoxifier genes might be subject to positive selection driven by changing substrates . To test this possibility , I used maximum-likelihood codon analysis on paralogous groups of genes from many parts of the overall CYP450 tree [14 , 15] . Most attention was given to tree regions where there was evidence of frequent recent gene duplication , both because this might indicate selection to diversify enzyme function and because this provided abundant sequence data for analysis . Specifically , 203 genes in clades with five or more closely-related genes were tested , including six groups from primates , nine from rodents , and six from a mixture of primates and rodents . Details of the method are given in Materials and Methods , and summary statistics for the results are given in Table S3 . A total of five of the 21 tested sets showed highly significant evidence of positive selection . The specific amino acid sites under positive selection were analyzed for a set of 12 rodent Cyp2d genes , which showed the strongest signal of positive selection . In this set , a Bayes-Empirical-Bayes method identified nine sites with strong evidence of positive selection ( p > 0 . 9 for each site ) . The structural positions of these nine sites were compared by alignment with two closely related CYP2 proteins for which crystal structures have been determined [16 , 17] . Both in the primary sequence alignment and in the 3-D protein structure , the nine sites under positive selection were clearly associated with substrate-binding or substrate-channel regions ( Figures S6 and S7 ) . In the protein structure , the distance from positive selection sites to a bound warfarin ligand was significantly smaller than for random sets of sites ( p ≈ 0 . 0002 , see Materials and Methods ) . In addition to these nine sites , all three small indels among the 12 rodent genes coincided with regions of substrate binding ( Figure S6 ) . No method exists to analyze such indels for positive selection , but it is plausible that they affect substrate interactions and might be subject to such selection . These results are consistent with changing xenobiotic substrates driving the observed positive selection . For comparison with the unstable genes , nine sets of genes from stable CYP450 groups were also tested for positive selection ( marked on Figure S1 ) . In each case , the orthologous genes from the six mammals were tested as a group . Because there are almost no gene duplications in these groups , it was not possible to include paralogs in these sets , but the degree of divergence of the genes ( total tree length ) was in the same range as for the paralog groups , indicating that the maximum-likelihood analysis should have similar power [18] . No evidence of positive selection was found for any of the nine sets , as expected based on their function as enzymes acting on endogenous substrates .
Phylogenetic instability among genes in families has been observed repeatedly , including among immunoglobulin genes [19] , major histocompatibility complex genes [20] , T-cell receptor genes [20 , 21] , mammalian olfactory receptor genes [22] , nematode chemosensory genes [15 , 23] , globin genes [24 , 25] , cichlid opsin genes [26 , 27] , zinc-finger genes [28] , F-box and MATH-BTB genes [29] , and others . Selective pressure from environmental change or pathogens has been suggested as a driver of duplication and diversification for many of these examples , but alternatives are difficult to rule out . The vertebrate CYP450 gene superfamily affords special insight into gene instability because probable biochemical functions of most of the genes are known and because it includes both stable and unstable genes . I found a strong correspondence between genes implicated in xenobiotic detoxification and evolutionary instability in the form of gene duplication and deletion . These findings strongly support the involvement of changing external selective pressure in high rates of birth-death gene evolution . Positive selection for amino acid change has been also been reported in many of the phylogenetically unstable gene families , suggesting that diversification is achieved by a combination of gene duplication and selection-driven divergence in sequence . For other unstable gene families , lack of stable family members for comparison or lack of extensive functional information prevent the highly informed analysis possible for the CYP450 genes . Nevertheless , a similar explanation for birth-death evolution is plausible for most or all such families . Immunoglobulin genes , major histocompatibility complex class I and class II genes , and T cell receptor genes function in pathogen defense , and an extensive literature documents gene duplications and deletions in these families during vertebrate evolution . Mammalian olfactory receptor genes and nematode chemosensory genes function in response to environmental chemical stimuli and might be subject to changes in desired ligand specificity or sensitivity . Lake Malawi cichlid fish have undergone duplications of opsin genes that have diverged to different spectral sensitivities , and expression of specific sets of these genes results in diverse adaptive visual properties , possibly driven by mate choice and foraging preferences [26 , 27] . In other cases the role of external selective forces in driving birth-death evolution is less obvious but plausible; for example the F-box and MATH-BTB genes in nematodes and plants have been hypothesized to function in pathogen defense , though the evidence is indirect [29] . Much less is known about specific functions of CYP450 genes in nematodes and plants . Based on maximum-likelihood protein trees with three Caenorhabditis species ( briggsae , elegans , and remanei ) and with the plants Arabidopsis thaliana and Oryza sativa ( rice ) , it is clear that birth-death evolution affects most members of the CYP450 family in these groups as well . For example , only 7% of Arabidopsis CYP450 genes and 38% of C . elegans genes had strict one-to-one orthologs within their groups ( unpublished data ) . As in the human and mouse genomes , closely related unstable genes tended to be in genomic clusters . The exact degrees of stability and instability cannot be directly compared because the phylogenetic distances differ , and genome annotation qualities may vary . A few CYP450 genes in humans lack clear functional information . The relative evolutionary stabilities of these genes may be predictive of general aspects of their functions . Among stable genes of unknown function , human CYP20A1 and orthologs are well separated from all other CYP450 genes on the protein tree and are encoded by a single gene in each of the ten species analyzed ( top left of Figure S1 ) . This pattern suggests an ancient and critical function that is shared among all the species . A less clear-cut but similar pattern is seen for an unnamed human CYP27 ( Chromosome 2 ) and the CYP2U1 and CYP2R1 genes . I speculate that these four enzymes will prove to be involved in biosynthesis or catabolism of endogenous substrates of importance throughout vertebrates . Consistent with this interpretation , CYP20A1 expression is enriched in immune system cell types , CYP2U1 is expressed broadly , and CYP2R1 expression is enriched in testis and immune system [30] . Unstable genes for which I could find no described function include CYP4A43 , CYP4Z1 , CYP4Z2 ( possible pseudogene ) , CYP4X1 , CYP4A11 , CYP4A22 , CYP4F2 , CYP4F3 , CYP4F8 , CYP4F11 , and CYP4F12 . Among these , in addition to birth-death evolution , the CYP4F group is subject to significant positive selection , consistent with substrate-driven selection for amino acid change . I speculate that each of these genes functions in xenobiotic detoxification or some other process that acts on environmental substrates . Consistent with this interpretation , expression of most of these genes is enriched in liver ( CYP4A11 , CYP4A22 , CYP4F2 , CYP4F3 , CYP4F11 , and CYP4F12 ) or in tracheal tissue ( CYP4X1 ) [30] , which are tissues where other xenobiotic CYP450 genes are known to function [6 , 7] . Finally , the evolution of the CYP2W1 and CYP2J2 genes is less clear , with duplications in some species but only a single gene in humans . CYP2W1 has a broad expression pattern [30] , and evolution of the group appears most consistent with a relatively stable function in mammals but extensive duplication and diversification in amphibians and teleost fish . CYP2J2 expression is enriched in liver [30] , and the group includes multiple genes in most species but only a single gene in humans and macaques . Groups of closely related unstable CYP450 genes are strongly clustered in the human and mouse genomes . A similar pattern of genome clustering is found in a number of other human and mouse gene families [31–33] . These patterns indicate that DNA duplications that persist in the genome occur predominantly locally . In contrast , about half of very recent large DNA duplications ( segmental duplications ) on the human lineage are interchromosomal , and some of the remainder occur nonlocally on the same chromosome [34–37] . It is possible that these segmental duplications represent a gene duplication mechanism that has dramatically and recently increased on the primate lineage [34] . Alternatively , nonlocal segmental duplications may be more unstable than local duplications and fail to persist over longer evolutionary periods , so that they rarely give rise to persistent functional duplicate genes . The combination of evolutionary instability and gene clustering has parallels with “genomic islands” found in bacteria . Genomic islands consist of blocks of DNA that are present in some bacterial isolates but absent from other closely related isolates ( reviewed in [38] ) . Though the full picture is still developing , it is likely that most genes in genomic islands have auxiliary organismal functions such as pathogenesis , antibiotic resistance , symbiosis , and specialized metabolism [38] . Despite some parallels , the mechanisms by which bacterial genomic islands and unstable gene islands in animals arise are different: clustering in bacteria is a consequence of horizontal transfer of contiguous blocks of DNA , whereas clustering in animals is restricted to closely-related genes and arises from local gene duplication events without horizontal gene transfer .
Sequence data were downloaded from ENSEMBL [39] release 40 , which contained the following genome builds: H . sapiens NCBI36 , M . mulatta MMUL1 , B . taurus Btau2 . 0 , C . familiaris BROADD1 , M . musculus NCBI36 , R . norvegicus RGSC3 . 4 , G . gallus WASHUC1 , X . tropicalis JGI4 . 1 , D . rerio ZFISH6 , and T . rubripes FUGU4 . 40 . Data for P . troglodytes were not included because the genes are too similar to human to be informative . Data for Monodelphis domestica and Tetraodon nigroviridis and several low coverage genomes were considered but rejected because current gene prediction sets appeared to be too incomplete or inaccurate ( unpublished data ) . Starting with a hand-collected set of human CYP450 sequences , all known isoforms of human CYP450 proteins were collected from a protein BLAST ( BLASTP ) search of all human proteins . Many genes had multiple transcript and protein forms . Each such case was analyzed to identify a single protein isoform that appeared full length and most family-typical , based on alignment with related CYP450 proteins . This analysis resulted in a set of 60 reference human proteins that were full length or nearly full length , including three possible pseudogenes . Proteins from other species were collected from BLASTP searches with these 60 human proteins as query . All isoforms were kept initially , and for each gene the isoform with the highest BLASTP score against a human protein was kept for further analysis ( for the most part , this approach discarded obviously incomplete proteins ) . Subsequent tree analysis of these proteins indicated the surprising absence of a few CYP450 proteins with endogenous substrates . Some of these were found to be present ( but unpredicted ) in the cognate genomes by TBLASTN searches ( Table S1 ) . Other genes are probably present in unassembled sequence; the number of missing genes and genome sequence coverage in the mammalian genomes is summarized in Table S2 . After discarding a few proteins that were missing more than 30% of the expected CYP450 protein or clearly aligned badly , the remaining set of 628 proteins were aligned with ClustalW ( default settings except gap clustering turned off ) [40] . A complete list of these protein sequences is found in Dataset S1 with ENSEMBL gene identifiers [39] . After removal of aligned columns with more than 30% gap residues , the resulting multiple alignment had 417 sites , at least 70% of which were present in any given sequence . A maximum-likelihood tree was constructed from this alignment with PhyML ( JTT matrix , four rate categories , gamma shape parameter 1 . 0 ) [41] . Tree nodes were rotated , colored , and displayed with Bonsai [42] . Additional annotations were added to the tree image for Figure S1 . The complete tree was subjected to 200 bootstrap repeats with PhyML with the same settings , which required ten days of central processing unit time . Subtrees of specific interest were tested with 1 , 000 bootstrap repeats ( Figure S2 ) . Multiple sets of five to 15 closely related primate and rodent genes were selected from the tree in Figure S1 . For each set , a protein alignment was made using ClustalW ( default settings except gap clustering turned off ) [40] . This protein alignment was used to generate the corresponding codon alignment and to construct a maximum-likelihood protein tree with PhyML [41] . The tree and codon alignment were analyzed with CodeML from PAML 3 . 14 [14] , using models 7 and 8 , with three starting dN/dS ( ω ) values for model 8 to avoid local optima . The neutral model 7 assumes a β-distribution of 10 dN/dS ratio classes constrained to lie between 0 and 1 . 0 , whereas the selection model 8 permits one additional dN/dS ratio class without constraint . In order to minimize the effects of gene prediction and alignment errors , aligned columns with a gap in any sequence were excluded from analysis ( “cleandata” option in CodeML ) . The transition/transversion ratio ( κ ) was estimated by CodeML . Statistical significance was assessed using a χ-square test on twice the difference in log-likelihood values ( ΔML ) for models 7 and 8 with two degrees of freedom , a statistic shown to be conservative in simulations [43] . Assignment of specific sites under likely positive selection was based on the Bayes-Empirical-Bayes test implemented in CodeML . In Figures S6 and S7 , marked sites are subject to positive selection with p > 0 . 9 . Extensive gene conversion can exaggerate signals of positive selection , but direct analysis of gene conversions among the tested genes suggests that they are unlikely to confound the analysis . GENECONV [44] was used to test for possible gene conversion events in the complete genomic regions of all the mouse and human genes analyzed for positive selection ( unpublished data ) . In all cases except one , detected regions of conversion affected only a few genes in <10% of their coding region , suggesting that they will not significantly affect the signal of positive selection . The exception was among mouse genes in the CYP2D family , but even in this case a minority of genes had evidence of gene conversion and none were affected in more than 20% of their coding region . The significance of the association of positive selection sites to bound warfarin ligand in the CYP2C9 structure was assessed as follows . The mean physical distance of the alpha backbone carbon atom to a reference warfarin atom was measured for the nine amino acid residues under positive selection . Warfarin atom C10 ( carbon 10 , atom 7439 of Protein Data Base structure 10G5 ) was used as the reference; it was selected as centrally located in the binding pocket/access channel based on visual inspection of the three-dimensional protein structure . The expected mean distance for randomly located amino acid residues was determined by repeatedly choosing nine random amino acid residues and computing the mean distance from their alpha backbone carbon atoms to the reference warfarin atom . Mean distances for 10 , 000 simulation repeats fitted a normal distribution ( Figure S8 ) , and all but two had mean distances larger than the mean distance for positively selected sites .
The Protein Data Base ( http://www . pdb . org ) number for the structure discussed in this paper is 1OG5 .
|
Genes vary greatly in their long-term phylogenetic stability , and there exists no general explanation for these differences . Stable genes persist as a single copy over a wide range of distantly related species , whereas unstable genes undergo frequent duplication and loss in a process called birth-death evolution . The vertebrate cytochrome P450 ( CYP450 ) gene superfamily includes many genes that are present in a single copy in species ranging from teleost fish to mammals and other groups of genes that undergo active birth-death evolution across the same species . The author found that nearly all stable CYP450 genes encode enzymes known to function in the synthesis and degradation of steroid and retinoid hormones ( and related molecules ) . These hormones function in core developmental pathways in vertebrates . In contrast , most unstable CYP450 genes encode enzymes that detoxify foreign small molecules ( called xenobiotics—foreign biochemicals ) . In addition , many of the unstable CYP450 genes are subject to natural selection to change their amino acid sequence over time ( positive selection ) , probably in response to changes in xenobiotic exposure . These findings suggest that stable and unstable genes differ in their rates of birth-death evolution , because stable genes have core endogenous functions whereas unstable genes respond to changing environmental conditions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] |
[
"teleost",
"fishes",
"chicken",
"vertebrates",
"evolutionary",
"biology",
"homo",
"(human)",
"genetics",
"and",
"genomics",
"mus",
"(mouse)"
] |
2007
|
Rapid Birth–Death Evolution Specific to Xenobiotic Cytochrome P450 Genes in Vertebrates
|
The yeast Candida albicans is a human commensal and opportunistic pathogen . Although both commensalism and pathogenesis depend on metabolic adaptation , the regulatory pathways that mediate metabolic processes in C . albicans are incompletely defined . For example , metabolic change is a major feature that distinguishes community growth of C . albicans in biofilms compared to suspension cultures , but how metabolic adaptation is functionally interfaced with the structural and gene regulatory changes that drive biofilm maturation remains to be fully understood . We show here that the RNA binding protein Puf3 regulates a posttranscriptional mRNA network in C . albicans that impacts on mitochondrial biogenesis , and provide the first functional data suggesting evolutionary rewiring of posttranscriptional gene regulation between the model yeast Saccharomyces cerevisiae and C . albicans . A proportion of the Puf3 mRNA network is differentially expressed in biofilms , and by using a mutant in the mRNA deadenylase CCR4 ( the enzyme recruited to mRNAs by Puf3 to control transcript stability ) we show that posttranscriptional regulation is important for mitochondrial regulation in biofilms . Inactivation of CCR4 or dis-regulation of mitochondrial activity led to altered biofilm structure and over-production of extracellular matrix material . The extracellular matrix is critical for antifungal resistance and immune evasion , and yet of all biofilm maturation pathways extracellular matrix biogenesis is the least understood . We propose a model in which the hypoxic biofilm environment is sensed by regulators such as Ccr4 to orchestrate metabolic adaptation , as well as the regulation of extracellular matrix production by impacting on the expression of matrix-related cell wall genes . Therefore metabolic changes in biofilms might be intimately linked to a key biofilm maturation mechanism that ultimately results in untreatable fungal disease .
Metabolism is a key driver of cell growth and division , and has a widespread influence on cell function . For example , metabolites can regulate gene expression [1] , metabolic enzymes can double as RNA binding proteins and regulate mRNA expression [2] , and the “power house of eukaryotic cells”–the mitochondrion–plays diverse roles in nuclear gene expression control [3] , as well as pathways of programmed cell death [4] and cellular aging [3 , 5] . We are studying mitochondrial functions in Candida albicans , a human commensal yeast known to cause serious infections in susceptible individuals [6] . Al Brown and colleagues have recently argued that metabolism should be put center stage for a holistic understanding of virulence and host interactions of human fungal pathogens [7] . C . albicans can inhabit several niches in the human body that differ in nutrient availability , and it has evolved sophisticated mechanisms to cope with changing nutrient environments . For example , C . albicans uses complex networks of transcriptional activators and repressors to modulate the switch from being a commensal inhabitant of the gastrointestinal ( GI ) tract , to becoming a pathogen localized in the blood stream [8 , 9] . Many of the target genes of these regulators relate to metabolic functions [8] . Like the majority of organisms , C . albicans is highly responsive to carbon source availability . Major metabolic remodeling , but also global changes in cell physiology including restructuring of the cell surface and host interactions , have been found when comparing C . albicans grown in the fermentative carbon source glucose with the non-fermentative carbon source lactate [10 , 11] . These carbon sources are found at varying concentrations in the GI tract , the vaginal tract and the bloodstream , and thus are relevant nutrients for C . albicans in host environments [7] . Metabolic control is linked to a critical virulence attribute of C . albicans—morphological plasticity , whereby this organism transitions between distinct cell types in response to environmental signals . One such developmental transition of central importance is substrate-attached growth of C . albicans in multicellular biofilm communities , a property that is highly relevant for virulence [12] . Biofilm formation involves several important phenotypic aspects , such as adherence , cell surface restructuring , the yeast-to-hyphae morphogenetic switch and the production of protective extracellular matrix material [13] . The pathways that drive adherence and morphogenesis have been widely studied , and several signal transduction pathways as well as a highly interconnected network of transcription factors are known to regulate biofilm formation in C . albicans [13 , 14] . Recent studies have begun to address the pathways required for extracellular matrix biogenesis ( i . e . making the matrix components ) [15 , 16] . However , the regulatory aspects of matrix production are poorly defined , and only two gene expression regulators are known to control matrix accumulation in biofilms: the transcription factor Rlm1 is a positive regulator [17] , while the transcription factor Zap1 is a negative regulator [18] . Transcriptomics and metabolomics analyses of biofilms have revealed that a critical difference between planktonic ( suspension ) growth and surface-attached biofilm growth relates to metabolic reprogramming . Glycolysis , ergosterol biosynthesis , the sulfur assimilation pathway , glycerol synthesis and respiratory metabolism are all modulated in biofilms [14 , 19–22] . Following from these studies , deletion of differentially expressed genes required for metabolic functions in biofilms has been found to impact on biofilm formation , underscoring the importance of metabolic reprogramming for the biofilm growth mode [23 , 24] . Relevant to our research interests , mitochondrial function and/or biogenesis are differentially regulated in biofilms [19] . These and other studies [25 , 26] are consistent with a role for mitochondrial reprogramming in biofilm formation . However , the important question of how metabolic changes are superimposed onto the developmental and structural changes that drive biofilm-dependent phenotypes remains poorly elucidated . Although mitochondrial activity is clearly important for fungal virulence [27] , very little is known about the regulatory networks that control mitochondrial biogenesis in C . albicans in an environmental and/or developmental context . Three transcription factors have been shown to have a role in mitochondrial respiratory activity in this pathogen [28] , and the transcriptional regulatory complex Mediator modulates respiratory metabolism , although the gene targets are not known [25] . In addition to these transcriptional mechanisms , we have previously shown that posttranscriptional regulation through the major cytoplasmic mRNA deadenylase complex Ccr4-NOT has an impact on mitochondrial biogenesis in C . albicans in planktonic cultures [29] . We report here that a proportion of the mitochondria-related genes down-regulated in mature C . albicans biofilms [14] belong to an mRNA network of putative targets of the Pumilio RNA binding protein Puf3 . We performed a detailed analysis of Puf3 function in the pathogen C . albicans , and our results suggest that posttranscriptional mRNA control of genes encoding the mitochondrial ribosome has been rewired between S . cerevisiae and C . albicans . We further show that posttranscriptional gene regulation is important for mitochondrial activity in biofilms , and demonstrate that that inactivation of the mRNA deadenylase CCR4 or uncoupling of mitochondrial oxidative phosphorylation have consequences for biofilm structure and maturation , particularly the production of the extracellular matrix . While secretion of the extracellular matrix is a property of biofilm growth that has very important consequences for biofilm resistance to antifungal drugs ( reviewed in [30] ) , and evasion of innate immunity [31] , it is the least understood of the biofilm maturation processes . Our study illuminates a new mechanism of biofilm matrix regulation , leading us to propose a model for how environmental and nutritional changes in biofilms drive a critical biofilm protection mechanism .
Inspection of the Nobile et al data [14] revealed that of the 622 genes down-regulated in biofilms relative to growth in suspension , 162 are annotated to the GO term “Mitochondrion” ( p = 1 . 40E-10 , FDR≈0 ) ( S1 Dataset and Fig 1A ) . Of these , based on inferred functions from homology to the model yeast S . cerevisiae , we judged that 29 genes are likely to have dominant functions in other organelles ( S1 Dataset ) . For example , some of the genes encode glycolytic enzymes , enzymes of ergosterol biosynthesis in the endoplasmic reticulum , cytoplasmic ribosome subunits and translation factors , and proteins with nuclear roles ( S1 Dataset ) . In several cases , their annotation to mitochondria is based on proteomics studies that found the proteins in the mitochondrial proteome [32 , 33] . However , caution has to be applied , particularly when abundant proteins , as well as structures associated with mitochondria , such as the endoplasmic reticulum and mitochondria-associated translation , are considered . Whether these proteins play a role in mitochondrial activity/biogenesis in C . albicans remains to be studied . The remaining 133 genes were sorted based on their putative roles into two groups: “activity” ( genes necessary for mitochondrial functions , such as metabolic enzymes ) ( 68 genes ) and “biogenesis” ( genes necessary for building the organelle ) ( 58 genes ) ( S1 Dataset; of note , 7 genes had unclear functions and were not assigned to either category ) . 36 of the 133 mitochondria-related genes , and about half of the “biogenesis” group ( 26/58 ) possess in their 3′ untranslated region ( 3′ UTR ) a predicted binding site for the RNA binding protein Puf3 as defined in S . cerevisiae: ( C/U/A ) ( A/G/C/U ) UGUA ( A/C/U ) AUA ( Fig 1A and S1 Dataset ) . Puf3 is a Pumilio family RNA binding protein which in S . cerevisiae controls mitochondrial biogenesis by impacting on the decay and subcellular localization to mitochondria of a network of mRNAs encoding mitochondrial proteins [34–37] , reviewed in [38] . Our analysis therefore suggests for the first time that gene regulation in C . albicans biofilms , and more specifically mitochondrial biogenesis , might be regulated by posttranscriptional mechanisms . The structure of the S . cerevisiae Puf3 RNA binding domain in complex with its cognate site from the COX17 3′ UTR has been solved [39] . Much like other PUF proteins , Puf3 uses the repeats that form the concave surface of its arc-shaped RNA binding domain to interact with the eight bases of the core RNA recognition motif . It discriminates its own targets from those bound by other yeast PUF family members by virtue of a binding pocket in repeat 8–8′ of the RNA binding domain that interacts with a 5′ cytosine at position -2 of the recognition element [39] ( Fig 1B ) . Although a cytosine at -2 leads to high affinity binding [39] , other nucleotides can be found at this position in Puf3 targets , as shown by the consensus sequence [34 , 35 , 40] . Primary sequence alignment showed conservation of the -2 cytosine-binding motif in repeat 8′ of the PUM domain of C . albicans Puf3 ( encoded by C4_05370W or orf19 . 1795 ) . ( Fig 1C ) . This suggests that C . albicans Puf3 recognizes the same motif as its S . cerevisiae homolog . To better define the Puf3-regulon in C . albicans , we performed a bioinformatics search for the ( C/U/A ) ( A/G/C/U ) UGUA ( A/C/U ) AUA recognition element in 3′ UTRs genome-wide . Firstly , we precisely defined the landscape of 3′ UTRs across the C . albicans transcriptome with a new 3′ sequencing technology that we developed called PAT-seq [41] . The 3' UTRs were called based on the most highly expressed peak within 400 bases of the end of the coding sequence , and not lying within a following gene on the same strand . Previous RNA-seq data has been informative in determining 3′ UTRs of C . albicans transcripts [42] , and our mapping correlates well with the study of Bruno et al ( Fig 2A ) . The apparent extension in length of 3′ UTRs in the Bruno et al data [42] is due to alternative adenylation that is present at lower abundance than the major peak called by our approach . The distinction between alternative 3′ UTR length isoforms is not easily extracted from regular RNA-seq , but is sensitively detected by PAT-seq . Moreover , with our technology we mapped 4862 3′ UTRs ( or 78% of the transcriptome ) , and could map an additional 2006 3′ UTRs , beyond the annotations published by Bruno et al ( S2 Dataset ) . Files to display 3′ UTR positions ( including alternate isoforms where they exist ) in CGD gbrowse are available ( see Materials and Methods ) . We performed an equivalent analysis in S . cerevisiae , where we could map 5402 3′ UTRs . Comparison of positions of adenylation with S . cerevisiae showed that in C . albicans 3′ UTRs are overall shorter . The offset between convergent and overlapping 3′ UTRs is also slightly shorter , with both of these features reflecting a higher level of compaction of the C . albicans genome ( Fig 2B ) . Despite this global similarity between 3′ UTR length distributions , the absolute 3′ UTR length in orthologous genes is not conserved between the two species ( Fig 2C ) . Analysis of gene ontology enrichment ( GO Function ) in the C . albicans dataset , in windows of 50 bases of 3′ UTR length and at p < 0 . 0001 , identified enrichment of very broad GO terms , such as “binding” , “anion binding” , “ATP binding” , “transferase activity” , “pyrophosphatase activity” , with one exception being “structural constituent of the ribosome” in the 50–99 bases 3′ UTR length group ( S3 Dataset ) . An equivalent analysis of the S . cerevisiae dataset similarly showed broad terms , with the exception being “electron carrier activity” that mapped to 3′ UTRs longer than 200 bases ( S3 Dataset ) . To further address if there is conservation of 3′ UTR lengths between the two yeasts related to gene function , we mapped genes with mitochondrial functions in Fig 2C ( shown as red dots ) . However , no correlation was seen in regards to 3′ UTR lengths for mitochondria-related transcript between the two yeast species . Armed with the 3′ UTR landscape for the C . albicans transcriptome , we identified a total of 555 putative Puf3 targets ( S2 Dataset ) . For comparative purposes , we searched the S . cerevisiae genome in an equivalent manner and identified 671 genes with Puf3 binding sites in their 3′ UTR ( S2 Dataset ) . In 3′ UTRs , the Puf3 motif occurs two or three times as often as random motifs of the same composition . However in the genomes as a whole Puf3-binding sequences are not more prevalent than random motifs of the same composition ( 2751 instances in S . cerevisiae , and 3463 instances in C . albicans ) . Of the 555 putative Puf3 targets in C . albicans , 432 ( 77 . 8% ) have an ortholog in S . cerevisiae ( S2 Dataset , Fig 3A ) . The number of genes where both species have the Puf3 motif is 198 ( Fig 3A ) . This correlation is highly significant by Fisher’s Exact Test , as similar analysis with a random shuffling of the motif generally produces a value less than 3 in common . Therefore , the Puf3 motif is much more highly conserved than shuffled motifs . The 3′ UTR length of the shared mRNAs with the Puf3 motif is not conserved ( S1 Fig , red dots ) . Furthermore , the position of the Puf3 binding site is not preserved relative to the stop codon or the polyadenylation site between the two species ( Fig 2D and 2E ) . Within the predicted Puf3 targets in C . albicans 268 genes or 48 . 3% are assigned to the GO term “Mitochondrion” ( p = 6 . 12e-141 , FDR≈0 ) ( S4 Dataset ) . With the same logic as described in the analysis of the biofilm-related genes , we judged that 12 of these genes encode proteins with predominant roles in another subcellular location . These are indicated in S4 Dataset , but for ease of comparison of the conserved and divergent putative Puf3 targets between C . albicans and S . cerevisiae , all analyses were performed with the 268-gene set . Inspection of the 198 conserved putative Puf3 targets between these two yeasts revealed that 177 belong to GO Mitochondrion ( p = 126e-121 , FDR≈0 , S2 Dataset ) . In other words , the vast majority of the 198 conserved putative Puf3 targets belong to the mitochondrial network , and there is little conservation outside of that ( Fig 3A ) . Cytosine at -2 is dominantly found in both C . albicans ( 115/198 ) and S . cerevisiae ( 127/198 ) , however less than half ( 84/198 mRNAs or ≈ 42 . 4% ) have a -2 C in both species ( S2 Dataset ) . The vast majority of genes in the C . albicans Puf3-dependent mitochondrial network encode functions required for organelle biogenesis , rather than metabolic functions , including almost the entire set of proteins that constitute the mitochondrial ribosome ( Fig 3B and S4 Dataset ) . Our results are consistent with a previous bioinformatic analysis that led to the proposition that Puf3 is an important regulator of mitochondrial biogenesis in the Saccharomycotina group of fungi [43] . Unlike the S . cerevisiae puf3Δ mutant , which shows reduced growth on non-fermentable carbon sources [34 , 43] , the C . albicans homozygous deletion mutant puf3Δ/Δ was able to grow as well as the wild type strain in all carbon sources tested: glucose , glycerol and lactate ( Fig 3C ) . Moreover , the mutant did not display any observable changes in mitochondrial morphology ( Fig 3D ) . However , a mitochondrial role for Puf3 in C . albicans was revealed under mitochondrial stress . The mutant was hypersensitive to carbonyl cyanide m-chlorophenylhydrazone ( CCCP ) , which uncouples electron transport through the respiratory chain from ATP synthesis ( Fig 3E ) . This phenotype was complemented by re-introduction of the wild type PUF3 gene into the mutant genome ( Fig 3E ) . PUF proteins negatively impact on mRNA stability by recruiting deadenylases such as Ccr4 , which digest the poly ( A ) tail and initiate decay , reviewed in [38] . Therefore , we next tested mRNA half-lives in the absence of Puf3 in C . albicans . Attempts to use the transcriptional inhibitor 1 , 10 phenantroline in C . albicans failed ( no repression of gene transcription was observed for a prolonged time , S2A Fig ) . Another transcriptional inhibitor , thiolutin , was effective in inhibiting transcription only at very high doses , and had a stabilizing effect on mRNA ( S2B and S2C Fig ) . To circumvent these technical problems with transcriptional inhibitors , we placed two Puf3 targets , MRPL25 encoding a mitochondrial ribosomal subunit , and COX23 encoding a cytochrome c oxidase assembly factor , under the repressible MET3 promoter . In this set up , transcription is “on” in the absence of methionine and cysteine in the medium , whereas addition of methionine and cysteine results in rapid repression . In these strains , only one copy of the Puf3-dependent gene is placed under the MET3 promoter , and therefore primers were used to specifically detect this copy and not the endogenous gene ( Fig 4A ) . This strategy was chosen over deleting the second copy because of the high impact of mitochondrial mutations on C . albicans fitness due to its petite negative nature [44] . Control experiments showed that the detection of the MET3p-controlled transcripts of MRPL25 and COX23 by quantitative PCR experiments was highly specific ( S3 Fig ) , and these strains had no observable growth defects in any of the carbon sources tested ( S4 Fig ) . To address how Puf3 impacts on the stability of its putative mRNA targets , the experiment was performed as follows: the gene was repressed initially , then transcription was induced for 10 minutes , followed by repression and monitoring of the decay of the newly synthesized mRNA ( see Materials and Methods ) . The experiments were done in glucose , and in the non-fermentable carbon source lactate , which is physiologically important for C . albicans in human host niches [7] . The half-lives of these transcripts in wild type cells were not affected by carbon source , as in both glucose and lactate the mRNAs were rapidly decayed with similar kinetics ( Fig 4B and 4C ) . During growth in either glucose or lactate , deletion of PUF3 resulted in slower decay of the MRPL25 and COX23 transcripts ( Fig 4B and 4C ) . We demonstrate that the effect of PUF3 is direct by showing that mutations of the Puf3 binding site in the 3′ UTR of COX23 ( in the core element or a -2 cytosine to alanine mutation ) , phenocopied the deletion of PUF3 in otherwise wild type strains ( Fig 4D ) . The result with the -2C to A mutation is consistent with an important function of -2C for Puf3 binding in C . albicans . Our results were somewhat surprising in light of previous publications in S . cerevisiae that showed that: a ) transcripts encoding mitochondrial proteins are stabilized during growth of a wild type strain in non-fermentable carbon sources compared to growth in glucose , and b ) Puf3 represses mRNA stability only in glucose , but not in non-fermentable carbon sources [36 , 45] . While previous studies in S . cerevisiae used several non-fermentable carbon sources , lactate was not directly tested . To test the effects of lactate in S . cerevisiae , we made use of a strain that carries an RNA polymerase II temperature sensitive mutation ( rpb1-1 ) which allows for repression of transcription at 37°C ( wild type and puf3 mutant , described in [36] ) . Similarly to C . albicans , in S . cerevisiae deletion of PUF3 had a stabilizing effect in both glucose and lactate on two transcripts that encode cytochrome c oxidase assembly factors: COX17 and COX23 ( Fig 5 ) . Moreover , in wild type cells the half-life was similar for these two transcripts in glucose and lactate media ( compare Fig 5A and 5B ) . The situation with the mRNA encoding the mitochondrial ribosomal subunit MRPL25 was different . Firstly , unlike in C . albicans , in S . cerevisiae the MRPL25 transcript was stabilized in lactate compared to glucose in wild type cells ( compare Fig 5A and 5B ) . Secondly , while deletion of PUF3 had a stabilizing effect in glucose ( albeit less pronounced than what is seen in C . albicans ) , in lactate the half-life for MRPL25 was the same in wild type and puf3Δ mutant cells ( Fig 5A and 5B ) . For two other transcripts encoding mitochondrial ribosomal proteins , MRP21 and MRPL11 , in wild type cells mRNAs decay was also fast in glucose media and slower in lactate , although stabilization in lactate was less pronounced than what was observed for MRPL25 ( Fig 5A and 5B , bottom two graphs ) . Deletion of PUF3 resulted in stabilization of MRP21 and MRPL11 in glucose ( Fig 5A ) , and also some stabilization , particularly for MRPL11 , was observable in lactate ( Fig 5B ) . As controls , we assayed three transcripts with mitochondrial functions , which do not contain Puf3 binding sites in the 3′ UTR: MMF1 , FUM1 and OAC1 . Deletion of PUF3 had no effect on the decay of these three transcripts in either glucose or lactate ( S5 Fig ) . Therefore , mRNAs that do not contain a Puf3 binding motif do not respond to deletion of PUF3 , suggesting that the observed stabilization of the COX and MRP genes in the puf3Δ mutant is specific . In S . cerevisiae cells grown in lactate the decay of all tested transcripts ( Puf3-dependent and Puf3-independent ) was altered compared to glucose . Moreover , it appeared that , after initial decay , a subpopulation of the transcripts remained stable over the course of the experiment ( see Fig 5 ) . This was not observed in C . albicans ( Fig 4 ) . Unlike in C . albicans , in S . cerevisiae transcription was inhibited by a shift to 37°C , and we wondered whether the combination of lactate media plus the temperature shift might be contributing to the nature of the decay . To test this , we repeated the experiment with the MET3p-MRPL25 strains of C . albicans in lactate , but shifting the strains to 37°C concomitant with addition of methionine and cysteine to the media to repress transcription . Indeed , in these conditions in C . albicans also we observed fast initial decay and a subpopulation of transcripts that remained stable over the course of the experiment ( Fig 4E ) . This result indicates that the combination of lactate and temperature shift results in altered mRNA decay for a subpopulation of transcripts . Importantly , consistent with what was observed when the experiment was done at room temperature , and in contrast to the result from S . cerevisiae , in C . albicans the half life of MRPL25 in the wild type remained very short in lactate at 37°C , and clear stabilization was seen in the puf3Δ/Δ mutant ( Fig 4E ) . A possible explanation for the altered decay curves observed in lactate at 37°C is the presence of two sub-populations of mRNAs: one that is decayed faster and one that is decayed slower through the use of alternative polyadenylation sites . This would be carbon-source dependent , as it is only seen in lactate and not in glucose ( Fig 5 ) . To test this , we used extension poly-A test ( ePAT ) [46] on three transcripts in S . cerevisiae , COX17 , and two other mRNAs that encode mitochondrial proteins: TOM70 , which contains a Puf3 binding site in its 3′ UTR and OM14 , which does not . No alternative 3′ UTRs were observed for COX17 in glucose or lactate media , arguing against the use of alternative polyadenylation sites in this transcript dependent on carbon source ( S6 Fig ) . However , both TOM70 and OM14 revealed a transcript with a shorter 3′ UTR that was stabilized in lactate media ( S6 Fig ) . Collectively , these results show that the use of polyadenylation sites can be influenced by carbon source , but not necessarily in relation to Puf3 function . As shown in Fig 1 , a proportion of the genes that contain a Puf3 binding site is differentially expressed in C . albicans biofilms . Therefore , we next investigated the roles of posttranscriptional gene regulation in biofilm formation . In addition to Puf3 ( this study ) , we have previously shown that the Ccr4 mRNA deadenylase is involved in mitochondrial function in C . albicans [29] . Puf3 and Ccr4 are functionally linked . Puf3 , like other PUF proteins , can bind to the Ccr4-NOT complex [47] , which is thought to be the mechanism by which Ccr4 is recruited to PUF-dependent mRNAs for poly ( A ) tail deadenylation and decay , reviewed in [38] . Consistent with a role for these posttranscriptional regulators in biofilm gene expression , deletion of either PUF3 or CCR4 resulted in differential expression of mitochondria-related genes in C . albicans biofilms ( Fig 6A ) . CCR4 had a more pronounced effect than PUF3 , as would be expected for a major mRNA deadenylase that is recruited to transcripts by multiple RNA binding proteins . Control genes that do not contain a Puf3 binding site in the 3′ UTR ( POR1 , MDM12 , MDM10 ) were not up-regulated in the puf3Δ/Δ mutant ( Fig 6A and S7 Fig ) , although POR1 was significantly up-regulated in the ccr4Δ/Δ mutant in line with a broader role for Ccr4 in gene expression ( Fig 6A and S7 Fig ) . The levels of HWP1 and TEF1 , which are not related to mitochondria , were not up-regulated in either mutant ( S7 Fig ) , arguing against non-specific higher levels of mRNAs due to lack of a major mRNA decay pathway . Consistent with milder effects on gene regulation , and suggestive of compensatory effects , the puf3Δ/Δ mutant formed biofilms of wild type structure ( S8 Fig ) . However , analysis of the ccr4Δ/Δ mutant revealed a clear role of posttranscriptional gene regulation in C . albicans biofilms . The biofilms formed by ccr4Δ/Δ showed altered structure in scanning electron micrographs , with an increase in yeast cells over filamentous cells and hyper-production of biofilm extracellular matrix material ( Fig 6B ) . The effect on hyper-production of extracellular matrix was observed in several biofilm growth media ( RPMI , Spider and YNB ) ( Fig 6B and S9 Fig ) , regardless of whether the mutant biofilm showed somewhat increased or decreased biomass ( in Spider and RPMI respectively , Fig 6C ) . Therefore , the changes to biofilm extracellular matrix of the ccr4Δ/Δ mutant are unlikely to be caused by changes in growth rates and final biomass . Consistent with mitochondrial dysfunction , ccr4Δ/Δ mutant biofilms returned very low levels of respiratory activity using the XTT-reduction assay , which depends on mitochondrial respiration ( Fig 6D ) . Similar biofilm phenotypes were displayed by pop2Δ/Δ , which is inactivated in the POP2 subunit of Ccr4-NOT that is also essential for mRNA deadenylase activity ( Fig 6B and S9 Fig ) . Quantification of extracted extracellular matrix relative to total biofilm biomass showed a ~2 fold increase in ccr4Δ/Δ mutant compared to wild type ( Fig 7A ) , and the ccr4Δ/Δ biofilm was less sensitive to treatment with zymolyase preparations ( Fig 7B ) . Zymolyase preparations contain 1 , 3 ß-glucanase , mannanase , protease and endochitinase activities [48] and are therefore likely to act on several components of the extracellular matrix [49] . Enzymatic activities that impact on glucan synthesis and remodeling have a major impact on biofilm matrix levels [15 , 16] , and our previous work showed that in the absence of CCR4 the relative levels of glucan in the cell wall are reduced [29] . Consistent with this , the level of matrix 1 , 3 ß-glucan as determined by the Glucantell kit was not higher in ccr4Δ/Δ biofilms ( Fig 7C ) . Furthermore , genes encoding enzymes required for glucan synthesis and remodeling , including BGL2 , XOG1 and PHR1 that have roles in 1 , 3 ß-glucan accumulation in the biofilm matrix [15] , were not significantly up-regulated in ccr4Δ/Δ mutant biofilms ( Fig 7D; we note that a trend toward higher levels of BGL2 was observed ) . We further assayed five genes that regulate the level of mannan in the biofilm extracellular matrix [16] . Of these , three ( ALG11 , MNN9 and VRG4 ) displayed significantly higher expression levels in ccr4Δ/Δ biofilms ( Fig 7D ) , suggesting that the activity of Ccr4 impacts on the expression of enzymatic activities that control biofilm matrix production . One of the main stresses experienced by biofilm cells is hypoxia [50 , 51] , and Ccr4 has been recently implicated in the response of C . albicans to hypoxia [52] . We therefore hypothesized that , additional to regulation of factors with roles in matrix carbohydrate accumulation , changes to mitochondrial activity in ccr4Δ/Δ biofilms , and potentially hypoxic adaptation might be responsible for the observed biofilm phenotypes of the mutant . Treatment with CCCP , which uncouples mitochondrial oxidative phosphorylation , mimics the early effects of hypoxia on the level of the genome-wide transcriptional response [52] . Therefore , we analyzed the structural features of biofilms in the presence of CCCP . For these experiments we predominantly used RPMI medium , because it supported biofilm growth significantly better than Spider medium upon mitochondrial inhibition . The dose of 20 μM CCCP is in line with previous studies in S . cerevisiae [53] , and was effective in causing mitochondrial stress: similarly to ccr4Δ/Δ biofilms , treatment with CCCP led to up-regulation of mRNAs with mitochondrial functions ( Fig 8A ) , which is a known consequence of mitochondrial perturbation [29 , 54] . CCCP had an effect on biofilm growth ( S10 Fig ) , but nevertheless a complex biofilm that structurally resembled the wild type formed ( Fig 8B ) . Similar to deletion of CCR4 , treatment of biofilms with CCCP led to the accumulation of extracellular matrix ( Fig 8B ) , and somewhat greater stability upon treatment with zymolyase ( Fig 8C ) . These results support the proposition that mitochondrial dysfunction triggers ECM accumulation in biofilms .
Following transcription , mRNAs can be organized into co-regulated networks termed “posttranscriptional operons” , through sharing of recognition elements bound by RNA binding proteins [55 , 56] . Compared to transcription , little is known about the function and evolution of posttranscriptional mRNA networks in either model or pathogenic fungi . We performed 3′-based RNA-seq that allowed us to map precisely 3′ UTRs on 4862 C . albicans transcripts , 1/3 of the transcriptome more than what has been previously defined [42] . We found no general conservation of 3′ UTR lengths between S . cerevisiae and C . albicans mRNAs . Also , while the regulation of the functional network of mitochondria-related mRNAs by Puf3 is conserved between S . cerevisiae and C . albicans ( this study and [43] ) , we show that the location of the Puf3 binding site relative to the stop codon or polyadenylation site is not conserved , arguing that these features might not be critical for Puf3-dependent regulation . This could mean a difference with another related PUF protein , Puf5 in S . cerevisiae , for which the position of the binding motif relative to the termination codon impacts on repression [57] . As in S . cerevisiae , in C . albicans cytosine at position -2 in the recognition element is prevalent in the mitochondria-related putative Puf3 targets , and our data shows that -2C is important for repression of COX23 ( Fig 4E ) . However , our analysis shows that the -2C is conserved in only about 50% of the mRNAs that contain a Puf3 recognition motif in both yeasts , suggesting it might not be as critical for Puf3-dependent regulation . Our data sheds light on the regulatory rewiring of gene expression in fungi . This has mostly been studied at the level of gene transcription , and information on the evolution of posttranscriptional mechanisms is scarce . Bioinformatic studies of the distribution of Puf3 and Puf4 binding motifs have suggested that their roles might have changed during fungal evolution [43 , 58] . However , biological insight into the regulation and relevance for organismal biology of these posttranscriptional networks in distinct species is lacking . A major functional group containing Puf3 binding sites is the mitochondrial ribosomal genes . In an elegant example of evolutionary change , genes encoding mitochondrial ribosomal proteins have been transcriptionally rewired between the Crabtree positive S . cerevisiae and the Crabtree negative C . albicans by loss of a promoter element in S . cerevisiae that enabled the uncoupling of the regulation of the cytoplasmic ribosome and rRNA biogenesis genes from that of the mitochondrial ribosome [59] . Our data shows differences in the regulation of mRNA stability of the mitochondrial ribosomal subunit MRPL25 between C . albicans and S . cerevisiae , suggesting that mitochondrial ribosomal gene expression has been rewired at multiple levels of control during fungal evolution . In glucose , MRPL25 was rapidly decayed in both yeasts . In lactate the decay of MRPL25 remained rapid in C . albicans , but it was slower in S . cerevisiae ( Figs 4 and 5 ) . Moreover , in C . albicans deletion of PUF3 caused stabilization of MRPL25 in both glucose and lactate , whereas in S . cerevisiae this was seen only in glucose . We could show that these differences in posttranscriptional regulation of MRPL25 by carbon source and Puf3 were not due to the use of a temperature shift to stop transcription in S . cerevisiae , as rapid decay and the effect of the puf3 mutation were maintained in C . albicans when the experiment included a temperature shift . We further considered the possibility that transcriptional up-regulation of MRPL25 in lactate media has caused a stoichiometry problem for the decay machinery in S . cerevisiae , thereby non-specifically slowing decay ( of note , in C . albicans the gene is regulated by the MET3 promoter and so it is dissociated from regulation by carbon source ) . However , we do not believe this to be the case: firstly , the induction of the COX genes and the MRPL25 gene in lactate versus glucose was uniform for all three genes and MRPL25 was induced the least ( ≈7 fold for COX17 , ≈5 fold for COX23 and ≈4 . 3 fold for MRPL25 ) ; yet only for MRPL25 we observed a change towards a longer half-life in lactate . Secondly , comparing transcript levels normalized to the loading control gene SCR1 in C . albicans and in S . cerevisiae at the time of transcriptional repression ( time point 0 ) showed similar relative levels in lactate media , and yet only in S . cerevisiae stabilization was observed . Different mitochondrial ribosomal subunit genes might be impacted to a different degree by carbon source and Puf3-dependent regulation: compared to MRPL25 , a smaller effect of carbon source on the half life was observed for MRP21 and MRPL11 , and the transcripts were stabilized in the puf3Δ mutant in glucose , and a smaller effect was seen in lactate . Collectively our data suggest that the differences in mRNA stability control for the mitochondrial ribosomal subunits between S . cerevisiae and C . albicans reflect regulatory rewiring to enable fine-tuning of mitochondrial biogenesis with metabolic status and cellular energy requirements . Data with MRPL25 suggest that this regulatory difference might be in part due to distinct control by carbon source over Puf3 activity in the two fungal species . A proportion of the mitochondria-related genes differentially expressed in C . albicans biofilms belongs to the Puf3-regulon ( S1 Dataset and Fig 1A ) . A role for posttranscriptional gene regulation in biofilm maturation was revealed here by studying the consequences of inactivation of the main mRNA deadenylase CCR4 , which mediates Puf3-dependent repression and has a dominant role in regulating mitochondrial activity in biofilms . While inactivation of both CCR4 and PUF3 results in higher levels of mitochondria-related genes , both of these factors have a positive role in mitochondrial biogenesis ( this study and [29] ) . As we have proposed , this could reflect the roles of mRNA targeting to mitochondria for localized translation , and translational inhibition at the mitochondrial surface to assist co-translational protein import into mitochondria [38] . Drastic structural alterations were observed in ccr4Δ/Δ biofilms , with morphogenetic change towards more yeast-form cells ( consistent with a role of Ccr4 in hyphal differentiation [29] ) , over-production of extracellular matrix and higher biofilm resistance to degrading enzymes . To our knowledge , Ccr4 is only the third gene expression regulator of biofilm matrix production in C . albicans , and the second negative regulator , the other one being the zinc-responsive transcription factor Zap1 [18] . Biofilms made by the zap1 mutant resemble those made by the ccr4 mutant in several aspects: they display hyper-production of extracellular matrix ( although the precise components of the matrix regulated by these two factors differ ) , and there are more yeast-form cells compared to the hyphae-rich wild type biofilms [18] . Our study therefore adds a key posttranscriptional gene expression regulator as an important factor determining biofilm maturation . To date , almost all of the genes known to affect C . albicans biofilm matrix relate functionally to cell wall integrity , reflecting the fact that cell wall carbohydrates β-glucans and mannans are important components of the matrix material [49] . In the case of Zap1 , effects on the expression of glucan hydrolyases and metabolism/quorum sensing via expression of alcohol dehydrogenases have been suggested to play a role in matrix regulation [18] . In contrast to Zap1 , Ccr4 does not regulate the levels of matrix 1 , 3 β-glucan , but instead several genes required for the accumulation of matrix mannan [16] were expressed at higher levels in biofilms of the ccr4Δ/Δ mutant . Given the prominent role of Ccr4 in both mitochondrial activity and cell wall integrity [29] , and a link between mitochondrial function and the cell wall in C . albicans and other fungi [27] , we further considered that modulation of mitochondrial function in biofilms due to hypoxia could be sensed as a stress signal , and direct the production of the protective extracellular matrix . Consistent with this idea , uncoupling of oxidative phosphorylation by CCCP , which mimics hypoxia [52] , caused hyper-production of extracellular material in biofilms and higher biofilm stability towards degrading enzymes . Ccr4 , while not absolutely required for the cells to execute a transcriptional response to hypoxia , had nevertheless a significant effect on a subset of genes regulated by hypoxia [52] . Our data indicates that post-transcriptional regulation of gene expression by Ccr4 might serve to adjust biofilm metabolism and maturation in response to the harsh hypoxic environment , by impacting on mitochondrial activity and on the expression of cell wall and matrix related genes . Consistent with this proposition , Sellam et al have shown that Ccr4 has a role in cell wall gene expression in hypoxia [52] . Based on the data presented here , we propose a model in which metabolic and mitochondrial reprogramming in biofilms drive the pathways of biofilm maturation ( Fig 9 ) . In this model , changes to mitochondrial activity and biogenesis in the biofilm environment , possibly due to hypoxia , constitute the signal that triggers activation of protective mechanisms . This ultimately leads to extracellular matrix accumulation , potentially through the cross-talk between mitochondrial function and the pathways of cell wall biogenesis and overall cell stability ( reviewed in [27] ) . Ccr4 is involved in the response to hypoxia , and it orchestrates biofilm maturation by adjusting the expression of cell wall genes with roles in matrix production , as well as by regulating mitochondrial biogenesis and activity . In conclusion , we suggest that the interface between metabolic and developmental restructuring in biofilms has important consequences for matrix production , a phenotype that is implicated in both antifungal and immune resistance of the biofilm growth mode . This should be considered in the context of antifungal strategies that target metabolic regulators .
The open reading frame ( ORF ) encoding the C . albicans Puf3 ( orf19 . 1795 , C4_05370W ) was deleted in the BWP17 strain background , using URA3 and ARG4 as selection markers . The complemented strain was made by cloning the PUF3 ORF plus promoter and terminator regions into the plasmid pDDB78 , and this plasmid was integrated into the mutant genome by directing it to the HIS1 locus . We noticed that the sequence of PUF3 in the BWP17 strain background somewhat differs from the sequence in the Candida Genome Database . There are deletions of 6 nucleotides at two positions: 1265–1270 and 1593–1598 in 2790 nucleotide long PUF3 ORF . This affects two repeats: two amino acids P and G are removed from a 4 PG repeat ( i . e . leaving 3 PG repeats: PGPGPG ) , and NN in a 9 N repeat , thereby converting it to a 7 N repeat ( NNNNNNN ) . These two deletions are outside of the PUM domain and do not affect Puf3 function , as shown by complementation of mutant phenotype in Fig 3 . To generate Candida MET3p repressible strains , one allele of COX23 and MRPL25 were placed under the control of the MET3 gene promoter by genomic integration via homologous recombination of the HIS1-MET3p cassette amplified from the template pFA–HIS1–MET3p plasmid provided by Jürgen Wendland [60] . To generate strains containing mutant forms of the Puf3-binding sites in the 3′ UTR of COX23 , we first synthesized the whole 2238 bp long cassettes of MET3p-COX23 ORF containing wild type , core and—C mutant 3′ UTR ( GenScript ) and cloned these cassettes in pDDB78 plasmid and integrated them into the HIS1 locus of wild type strain of BWP17 background . All primers used for strain construction are listed in S1 Table . The C . albicans ccr4 and pop2 mutant strains were described in [29] . For most experiments the wild type control was strain DAY185 , which is a fully prototrophic strain derived from BWP17 by re-integration of the URA3 , ARG4 and HIS1 markers . The S . cerevisiae puf3Δ strain in the rpb1-1 background and the corresponding control strain were gifts from Wendy Olivas and are described in [36] . Generally , strains were grown in YPD ( 1% yeast extract , 2% peptone , 2% glucose ) medium , supplemented with 80 μg/ml of uridine for C . albicans . Standard growth temperature was 30°C . For analysis of sensitivity of puf3Δ/Δ on different carbon source , ten-fold serially diluted cultures of all strains , were spotted on YP plates containing 2% glucose ( YPD ) , 3% glycerol ( YPG ) or 3% lactate ( YPL ) and photographed after 2 days of growth . For sensitivity to CCCP , ten-fold serial dilutions of cultures of wild type and mutant strain were spotted on YPG plates with or without CCCP at the concentrations indicated in Fig 3 . For mRNA decay experiments of Candida strains ( Fig 4 ) , wild type and puf3Δ/Δ strains were first grown overnight and then diluted to OD600 = 0 . 1 and grown further for 4–5 hours in non-inducing YPD or YPL medium at 30°C . To induce expression from MET3p , the cultures were washed once with PBS and transferred into synthetic medium lacking methionine and cysteine followed by incubation for 10 min at 30°C . After incubation , the time = 0 min sample was collected , then methionine and cysteine were added to the remaining cultures to repress transcription from MET3p ( 2 . 5 mM methionine and 1mM cysteine ) . Samples were collected at indicated time points , flash frozen on dry ice and stored in -80°C until ready for RNA isolation . The experiment was performed 3–4 independent times , assaying one wild type and one mutant culture each time . The experiments with the mutated Puf3 binding site in the COX23 3′ UTR were performed twice independently . For decay experiment of S . cerevisiae strains ( Fig 5 ) , wild type and puf3Δ mutant were grown overnight in YPD or YPL medium at 24°C and next day cultures were diluted and grown for few hours until log phase ( OD600 ~ 0 . 8 ) . To repress transcription , cultures were quickly centrifuged , and then transferred to medium preheated to 37°C . Samples were collected at indicated time points , flash frozen on dry ice and stored in -80°C prior to RNA isolation . For gene expression analysis in C . albicans biofilms ( Figs 6A , 7D and 8A ) , biofilms were grown in 6 well microtiter plates in Spider medium for 48 hours in a 37°C incubator with shaking at 75 rpm . Cell pellets were harvested and stored at -80°C until use . RPMI 1640 buffered with MOPS ( RPMI-MOPS , pH = 7 . 2 , for a detailed formula refer to CLSI guideline M27-A3 ) and Spider medium ( 1% nutrient broth , 1% D-mannitol , 2 g K2HPO4 , pH = 7 . 2 ) were used for C . albicans biofilm growth . Biofilm growth in YNB media was performed essentially as described [61] . Mitochondrial morphology in Fig 3D was analyzed by staining log-phase cultures of the indicated C . albicans strains with 1 μM MitoTracker Red CMXRos ( Life Technologies , M7512 ) for 30 min in the dark at 30°C . Cells were washed with fresh medium , mounted on glass slides and observed with an Olympus BX60 fluorescence microscope . Images were taken with a 100x objective and analyzed using Spot Advanced Software ( http://www . spotimaging . com/software/ ) . For quantitative PCR ( qPCR ) analysis of all samples , total RNAs were extracted using the hot-phenol method followed by DNase I ( Ambion ) treatment to remove contaminating genomic DNA . Reverse transcription reaction was performed with Superscript III ( Invitrogen ) using 800 ng of total RNA and 200 ng of mammalian RNA as a “spike-in” control . qPCR was performed using the Fast-Start universal SYBR Green Master ( Roche ) on the LightCycler 480 ( Roche ) . The expression levels of the transcripts were normalized to the level of the SCR1 gene transcribed by RNA polymerase III . Analysis of qPCR data was performed using LinReg [62 , 63] . The primers used for qPCR analysis are listed in S2 Table . The ePAT assays were performed using RNA extracted from S . cerevisiae cultures grown in either glucose or lactate as the carbon source . The assay was performed as described in [46] . Primers are shown in S2 Table . Biofilm formation of C . albicans wild type and mutant strains was quantitatively assessed by crystal violet staining , as described in our previous study [64] . The growth medium was Spider or RPMI in 96-well microplates at 37°C . Following 90 min adhesion and washing of non-adherent cells , biofilms were grown for 48 h , with fresh medium added at 24 h . Negative control were wells contained medium-only . Alternatively , C . albicans biofilms were grown in 6 well plates under the same conditions and total biofilm biomass from three wells for each strain was determined by dry weight measurement . For the XTT reduction assay in Fig 6D , biofilms were grown in RPMI-MOPS for 48 h in 96-well plates , XTT added at 0 . 5 mg/ml , and reading made at OD492 following incubation in the dark for 2 h . Scanning electron microscopy was used for biofilm qualitative studies essentially as previously described [64] . Biofilms were formed on serum-coated silicone disks and grown for 48 h at 37°C with shaking at 75 rpm . Samples were viewed under a Hitachi S570 scanning electron microscope . Biofilms were grown with RPMI-MOPS in 96 well plates for 24 hours , and zymolyase assays performed essentially as described [15] , with the exception that crystal violet staining was used for quantification . Zymolyase–20T , MP Biomedicals was used . Individual C . albicans biofilm components , including biofilm cells and extracellular polymer matrix ( ECM ) , were isolated from 48 h mature biofilms grown in Spider medium , following the method by Taff et al [15] . Total ECM was then extracted and quantified relative to total biofilm consisting of cells and extracellular material using a method modified from McCourtie and Douglas [65] . β-1 , 3 glucan in the biofilm ECM was determined quantitatively using the Glucatell Endpoint Kit ( DKSH ) . Statistical analysis was performed on the biological repeats . Where multiple technical repeats were performed for a given biological repeat , the values of the technical repeats were averaged to give the data point for that particular biological sample . Statistical significance was calculated in GraphPad Prism using the Student’s t-test . The PAT-seq data utilized for 3′ analysis here is available in the NCBI Short Read Archive ( accession number SRP056994 ) . The tracks suitable for upload of all adenylation-sites we identified in the C . albicans transcriptome into gbrowse ( CGD ) the follow steps can be taken: 1 ) open gbrowse and set genome data source to A21 assembly; 2 ) go to ‘select custom tracks’; 3 ) copy and paste the web link within verma-gaur et al . , PAS-5 . wig; 4 ) go back to browser and click on the little spanner symbol to change the track-height to 80 . Note , use the down cursor rather than mouse; 5 ) Do the same for forward and reverse tracks . The S . cerevisiae data is drawn from [41] , and is available from GEO accession GSE53461 and for interactive viewing at http://rnasystems . erc . monash . edu/ . For PAT-seq the C . albicans strain was grown in YPD+Uridine to mid log phase at 30°C , and the PAT-seq experiment was performed exactly as previously described [41] , except that a template oligonucleotide compatible with SOLiD sequencing [/5BioTEG/CTGCTGTACGGCCAAGGCGTTTTTTTTTTTT] was used to append the 3’ tag , and SOLiD compatible 5’ linkers were ligated to the 5’ end . PAT-seq cDNA was input into 16 cycles of amplification with a SOLiD universal sequencing primer [CCACTACGCCTCCGCTTTCCTCTCTATGGGCAGTCGGAGAT] and SOLiD Barcoding primers ( Life Technologies ) . PAT-seq libraries were sequenced on a SOLiD 5500xl instrument according to the manufacturer’s instructions at the Gandel Charitable Trust Sequencing Centre ( Monash University ) . The data were mapped to the reference genome: C . albicans SC5314 assembly 21; using the tail-tools pipeline version 0 . 31 and nesoni version 0 . 117 ) http://rnasystems . erc . monash . edu/ . Figures were generated in R and Illustrator . The Venn diagram in Fig 3 was generated using US DOE Venn Diagram Plotter software ( http://omics . pnl . gov/software/venn-diagram-plotter ) , and we acknowledge PNNL and the OMICS . PNL . GOV website . For comparative analyses , the list of S . cerevisiae and C . albicans orthologs was obtained from the Candida Genome Database ( candidagenome . org ) . For the analysis of the relationship between 3′ UTR length and mitochondrial function , the list of genes annotated to GO mitochondria was from Amigo 2 , GO:0005739 .
|
Metabolism is a master regulator of cell biology , including gene regulation , developmental switches and cellular life-death decisions , with the mitochondrion playing a central role in eukaryotes . For the yeast Candida albicans mitochondrial functions have been implicated in host-pathogen interactions , but the regulatory mechanism that control mitochondrial biogenesis are poorly described . We identified the RNA binding protein Puf3 as a new mitochondrial regulator in C . albicans , and show that posttranscriptional regulation and mitochondrial function have important roles during community growth in biofilms . Perturbation of mitochondrial activity or inactivation of a key posttranscriptional regulator , CCR4 , led to changes in biofilm maturation , shedding light on the interface between metabolic reprogramming and biofilm developmental pathways . We illuminate a new mechanism that regulates extracellular matrix production , an essential biofilm feature that mediates the notorious drug resistance and immune evasion properties of the biofilm growth mode .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Integration of Posttranscriptional Gene Networks into Metabolic Adaptation and Biofilm Maturation in Candida albicans
|
Rapid advances in sequencing technologies set the stage for the large-scale medical sequencing efforts to be performed in the near future , with the goal of assessing the importance of rare variants in complex diseases . The discovery of new disease susceptibility genes requires powerful statistical methods for rare variant analysis . The low frequency and the expected large number of such variants pose great difficulties for the analysis of these data . We propose here a robust and powerful testing strategy to study the role rare variants may play in affecting susceptibility to complex traits . The strategy is based on assessing whether rare variants in a genetic region collectively occur at significantly higher frequencies in cases compared with controls ( or vice versa ) . A main feature of the proposed methodology is that , although it is an overall test assessing a possibly large number of rare variants simultaneously , the disease variants can be both protective and risk variants , with moderate decreases in statistical power when both types of variants are present . Using simulations , we show that this approach can be powerful under complex and general disease models , as well as in larger genetic regions where the proportion of disease susceptibility variants may be small . Comparisons with previously published tests on simulated data show that the proposed approach can have better power than the existing methods . An application to a recently published study on Type-1 Diabetes finds rare variants in gene IFIH1 to be protective against Type-1 Diabetes .
Common diseases such as diabetes , heart disease , schizophrenia , etc . , are likely caused by a complex interplay among many genes and environmental factors . At any single disease locus allelic heterogeneity is expected , i . e . , there may be multiple , different susceptibility mutations at the locus conferring risk in different individuals [1] . Common and rare variants could both be important contributors to disease risk . Thus far , in a first attempt to find disease susceptibility loci , most research has focused on the discovery of common susceptibility variants . This effort has been helped by the widespread availability of genome-wide arrays providing almost complete genomic coverage for common variants . The genome-wide association studies performed so far have led to the discovery of many common variants reproducibly associated with various complex traits , showing that common variants can indeed affect risk to common diseases [2] , [3] . However , the estimated effect sizes for these variants are small ( most odds ratios are below ) , with only a small fraction of trait heritability explained by these variants [4] . For example , at least loci have been identified for height , but these loci together explain only of the estimated heritability for this trait [5] . One possible explanation for this missing heritability is that , in addition to common variants , rare variants are also important . Evidence to support a potential role for rare variants in complex traits comes from both empirical and theoretical studies . There is an increasing number of recent studies on obesity , autism , schizophrenia , epilepsy , hypertension , HDL cholesterol , some cancers , Type-1 diabetes etc . [6]–[15] that implicate rare variants ( both single position variants and structural variants ) in these traits . From a theoretical point of view , population genetics theory predicts that most disease loci do not have susceptibility alleles at intermediate frequencies [16] , [17] . With rapid advances in next-generation sequencing technologies it is becoming increasingly feasible to efficiently sequence large number of individuals genome-wide , allowing for the first time a systematic assessment of the role rare variants may play in influencing risk to complex diseases [18]–[21] . The analysis of the resulting rare genetic variation poses many statistical challenges . Due to the low frequencies of rare disease variants ( as low as , and maybe lower ) and the large number of rare variants in the genome , studies with realistic sample sizes will have low power to detect such loci one at a time , the way we have done in order to find common susceptibility variants [5] , [22] . It is then necessary to perform an overall test for all rare variants in a gene or , more generally a candidate region , under the expectation that cases with disease are different with respect to rare variants compared with control individuals . Several methods along these lines have already been proposed . One of the first statistical methods proposed for the analysis of rare variants [23] is based on testing whether the proportion of carriers of rare variants is significantly different between cases and controls . A subsequent paper by Madsen and Browning [24] introduced the concept of weighting variants according to their estimated frequencies in controls , so that less frequent variants are given higher weight compared with more common variants . Price et al . [25] extended the weighted-sum approach in [24] to weight variants according to externally-defined weights , such as the probability of a variant to be functional . One potential drawback for these methods is that they are very sensitive to the presence of protective and risk variants . We introduce here a new testing strategy , which we call replication-based strategy , and which is based on a weighted-sum statistic , but that has the advantage of being less sensitive to the presence of both risk and protective variants in a genetic region of interest . We illustrate the proposed approach on simulated data , and a real sequence dataset on Type-1 diabetes .
If external information is available on the plausibility of a rare variant to be related to disease , it is of interest to be able to incorporate such information into our testing strategy . Such information has proved essential in the mapping of the disease genes for two monogenic disorders [26] , and may well prove important for mapping disease genes in more complex diseases . It is straightforward to extend the proposed approach to take into account such information . If we denote by the probability that a variant is functional , then we can rewrite the statistic above as:where signifies that variant occurs times in controls , and times in cases . In particular , if for all variants then we recover the statistic S above , where functional information was not used . If on the other hand a variant is not functional , then , and this variant is ignored .
We also applied our approach to a dataset on Type 1 Diabetes ( T1D ) , published by Nejentsev et al . [15] . In their paper , the authors resequenced exons and splice sites of ten candidate genes in cases and controls ( more details on the dataset are in Text S2 ) . In their study , rare variants were tested individually , and two SNVs in gene IFIH1 and two other SNVs in gene CLEC16A were found to be protective against T1D . Here we reanalyze the dataset using the proposed approach , and two of the existing approaches . For each gene and each method , we perform two-sided tests , testing for the presence of risk or protective variants . Results are in Table 5 . As in [15] we found one gene , IFIH1 , to be significant with all three methods ( P-value for all three methods ) . For this gene , controls were enriched for rare mutations compared with cases . Some evidence of enrichment in protective variants was also observed in another gene , CLEC16A , although the P-values do not remain significant after multiple testing correction .
We have proposed here a new testing strategy to examine associations between rare variants and complex traits . The approach is based on a weighted-sum statistic that makes efficient use of the information the data provides on the presence of disease variants in the region being investigated . The proposed test is based on computing two one-sided statistics , designed to quantify enrichment in risk variants , and protective variants , respectively . This aspect allows the proposed approach to have substantially better power than existing approaches in the presence of both risk and protective variants in a region . Even when only one kind of variants is present , we have shown via simulations that the proposed approach has consistently better power than existing approaches . An application to a previously published dataset on Type-1 Diabetes [15] confirmed the original finding , namely that rare variants in IFIH1 confer protection towards disease . The weights underlying our weighted-sum statistic depend only on the data at hand . However , external information on the likelihood of a variant to be functional could prove very useful , and could be combined with the information present in the data to improve power to identify disease susceptibility variants . Such information has been successfully used to identify the genes for several monogenic disorders [26] . Price et al . [25] discuss a weighted-sum approach with externally-derived weights , and show that such information can be very useful using several empirical datasets . We have also described a natural way to take into account such external functional predictions within the proposed framework . Since empirical data are only now becoming available , it is not known how often both risk and protective variants are present in a particular disease gene . When both types of variants are present , it seems appealing to be able to combine the two types of signals . It is possible to extend the proposed approach to take advantage of both kinds of disease variants , and we discuss such an extension in Text S4 . We noticed in our simulation experiments that such a hybrid approach can have much improved power when both types of variants are present , but this comes at the price of reduced power when only one type of variants is present . Therefore , depending on the underlying disease model , both approaches could provide useful information . The proposed approach is applicable to a case-control design and therefore is susceptible to spurious findings due to population stratification . Population stratification has been shown to be an important issue in the context of common variants . For rare variants , differences in rare variant frequencies between populations are likely to be even more pronounced . Development of new methods , and extension of existing methods are necessary to adequately address the issue . Alternatively , family-based designs offer the advantage of being robust to false positive findings due to population stratification . Replication of association signals in independent datasets is an essential aspect of any disease association study , and has become standard practice for common variants . Rare variants , due to their low frequencies and potential modest effects , are normally tested together with other rare variants in the same unit , e . g . , gene . Therefore a reasonable first replication strategy is at the level of the gene . Follow-up of individual variants in the gene can be performed to investigate whether any of the rare variants in the gene can be found to be significantly associated with disease . Finding rare disease susceptibility variants is a challenging problem , especially due to their low frequencies and the probable moderate effects on disease . So far the methods proposed in the literature have focused on case-control designs . However , for rare variants , family-based designs may prove very useful . Not only are they robust against population stratification , but they may also offer increased power due to the increased likelihood of affected relatives to share the same rare disease variants . Continued development of novel statistical methods for identifying rare disease susceptibility variants is needed for population-based designs , and especially for family-based designs . Software implementing these methods is available at: http://www . mailman . columbia . edu/our-faculty/profile ? uni=ii2135 .
|
Risk to common diseases , such as diabetes , heart disease , etc . , is influenced by a complex interaction among genetic and environmental factors . Most of the disease-association studies conducted so far have focused on common variants , widely available on genotyping platforms . However , recent advances in sequencing technologies pave the way for large-scale medical sequencing studies with the goal of elucidating the role rare variants may play in affecting susceptibility to complex traits . The large number of rare variants and their low frequencies pose great challenges for the analysis of these data . We present here a novel testing strategy , based on a weighted-sum statistic , that is less sensitive than existing methods to the presence of both risk and protective variants in the genetic region under investigation . We show applications to simulated data and to a real dataset on Type-1 Diabetes .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/complex",
"traits",
"mathematics/statistics"
] |
2011
|
A New Testing Strategy to Identify Rare Variants with Either Risk or Protective Effect on Disease
|
Understanding the relationship between genetic variation and gene expression is a central question in genetics . With the availability of data from high-throughput technologies such as ChIP-Chip , expression , and genotyping arrays , we can begin to not only identify associations but to understand how genetic variations perturb the underlying transcription regulatory networks to induce differential gene expression . In this study , we describe a simple model of transcription regulation where the expression of a gene is completely characterized by two properties: the concentrations and promoter affinities of active transcription factors . We devise a method that extends Network Component Analysis ( NCA ) to determine how genetic variations in the form of single nucleotide polymorphisms ( SNPs ) perturb these two properties . Applying our method to a segregating population of Saccharomyces cerevisiae , we found statistically significant examples of trans-acting SNPs located in regulatory hotspots that perturb transcription factor concentrations and affinities for target promoters to cause global differential expression and cis-acting genetic variations that perturb the promoter affinities of transcription factors on a single gene to cause local differential expression . Although many genetic variations linked to gene expressions have been identified , it is not clear how they perturb the underlying regulatory networks that govern gene expression . Our work begins to fill this void by showing that many genetic variations affect the concentrations of active transcription factors in a cell and their affinities for target promoters . Understanding the effects of these perturbations can help us to paint a more complete picture of the complex landscape of transcription regulation . The software package implementing the algorithms discussed in this work is available as a MATLAB package upon request .
With advances in whole genome high-throughput technologies such as ChIP-Chip , expression , and genotyping arrays , it is now possible to integrate data from these sources together to decipher the complex regulatory networks that govern transcription . In addition to serving as powerful models for how basic cellular function is achieved , these regulatory networks can also help us shed light on how certain disease phenotypes are manifested . At the heart of these networks are a few regulator genes such as transcription factors ( TFs ) , miRNAs and histones whose activity govern the behavior of many other genes . Among these regulators , transcription factors that bind the promoter regions of genes are by far the most well understood . The process of TFs activating or repressing transcription at initiation is believed to be the primary mechanism of gene regulation . A central question in genetics is how genetic variations perturb this underlying regulatory mechanism to give rise to differential gene expression and ultimately complex phenotypes . The simplest analysis one can perform to address this question is expression quantitative trait loci ( eQTL ) mapping , which identifies genetic variations such as SNPs in the form of linkages and associations that are correlated with gene expression . Such studies have been carried out in a variety of organisms including yeast [1] , [2] Arabidopsis [3] , mouse [4] , [5] and human [6]–[8] . These studies have identified many linkages between SNPs and genes in close proximity suggesting potential local regulatory mechanisms mediated by regulators such as transcription factors and miRNAs . These studies have also identified a few SNPs linked to the expressions of many genes suggesting a global regulatory mechanism mediated by master regulators such as transcription factors and histones . Unfortunately , beyond nominating candidate genes either as targets or regulators , these studies give little insight into how SNPs perturb the underlying transcription regulatory networks that control gene expression . To gain a better understanding of the mechanisms of transcription regulation , several systems biology based methods have been proposed including clustering of co-regulated genes [9] , multipoint linkage analysis [10] , [11] , pathway enrichment analysis [12]–[16] , prediction of regulatory modules [17] , [18] and the prediction of causal regulatory relationships [19]–[23] . Many of these advanced methods aim to tease out both the nodes ( regulators and targets ) as well as the topology ( mapping of edges ) in a transcription regulatory network from only considering gene expression profiles . Although these methods have predicted some interesting relationships , there are at least two aspects of transcription regulation that go unaddressed when we use them to study transcription factors and their targets . First , most previous methods rely on probabilistic models that do not provide much insight into the hidden dynamics between the activity of transcription factors and the expression of their targets . Second , the relationships inferred by these methods from the expression profiles alone can be misleading because the in vivo activity of a transcription factor does not always correlate with its expression levels [24] , [25] . To overcome these problems , we adopt a framework from network component analysis ( NCA ) [26] that considers a simple bipartite network model of transcription regulation involving only transcription factors and their targets . In this model , the expression of a target gene is completely captured by two properties of the network , the concentrations and promoter affinities of transcription factors . In general , inferring these two quantities from the expression profiles of the target genes alone is difficult . But by leveraging protein-DNA binding data from ChIP-Chip experiments [27] , [28] , a partial topology of the network can be constructed and one can make the inference given certain constraints [26] . The NCA method as described by liao et al . has been successfully applied to several gene expression datasets to understand transcription regulation in a temporal setting [26] and in the context of gene knockouts [29] . In this study , we extended NCA to study transcription regulation over a population gradient by modeling three mechanisms by which genetic variations perturb the concentrations and promoter affinities of active transcription factors to induce differential expression . Figure 1 gives a simple example that illustrates the original NCA model and our extensions . Imagine we have a small experiment where we collected the gene expressions of four genes , the genotypes of three markers over three individuals . Given the topology of the bipartite network between transcription factors and their targets ( Figure 1B ) , the NCA algorithm allows us to infer the active transcription factor concentrations ( C ) and the respective promoter affinities ( PA ) from the given gene expressions ( E ) in a log-linear fashion ( Figure 1A , see Methods ) . In this example , SNP1 and SNP3 are linked to the expressions of G1 and G3 while SNP2 is linked to the expressions of G2 and G4 . We propose three possible mechanisms any one SNP can perturb the regulatory network and show an instance of each using the given example . Because the inclusion of genetic variation creates additional parameters in each of our three models compared to the original NCA model , we expected them to always fit the data better . To effectively evaluate our models , we devised a likelihood ratio statistic and a permutation scheme to assess the statistical significance of our improvements . We then applied our method to study an expression data collected over 112 segregants of Saccharomyces cerevisiae yeast and two separate ChIP-Chip datasets generated by Harbisonet al . and Lee et al . . We identified several interesting global regulatory networks perturbed by SNPs located in regulatory hotspots . Some of these networks have one property perturbed ( transcription factor concentration or promoter affinity ) while others have both properties perturbed suggesting a complex mechanism of global regulation . We also examined linkages between SNPs and target genes located in close proximity . We found that many of these cis linked SNPs perturb the promoter affinities of transcription factors on a target gene locally confirming previous hypotheses of cis regulation . An interesting method proposed by Sun et al . also used the NCA framework to infer the concentrations of active transcription factors from gene expression data collected over the same yeast strains . Their method was designed to detect linkages between the inferred concentrations and genetic variations and used conditional independence tests to find modules of genes controlled by the same causal regulator . Compared to this method , we expect to find similar networks of genes and transcription factors but our method does not allow us to infer additional causal relationships using statistical tests . Instead , we focus on identifying different mechanisms by which genetic variations can perturb the regulatory networks by directly modeling the effects of these perturbations into the NCA framework . We do not attempt to make rigorous causal claims but use the causal information inherent in genotyping and ChIP-Chip experiments to suggest possible mechanisms of transcription regulation .
The NCA framework is a natural model for describing how transcription factors regulate gene expression . At the heart of the model is a log linear equation that relates the expression levels of genes collected over a gradient ( E ) to the concentrations ( C ) and promoter affinities ( PA ) of active transcription factors . Such a model is well supported by known kinetic properties of protein-DNA interactions [30] . In linear model terms , the transcription factor concentrations are the regressors , the gene expression levels are the response variables and the promoter affinities are the coefficients that relate the two . Figure 2B shows the log-linear equations describing the graph shown in Figure 1B . The goal of NCA is to infer the matrices of concentrations and promoter affinities from the matrix of gene expressions under some restrictions in the least squares sense . Treating genetic differences between individuals as a gradient , we applied this model to infer the matrices and from gene expressions collected from a population of yeast strains , . For the inference to have been possible , we removed a number of transcription factors and target genes to construct a network from the original ChIP-Chip data that met certain constraints [26] . After preprocessing the Lee et al . ChIP-Chip dataset , we were left with a network with 100 transcription factors and 2 , 294 target genes . Similarly , preprocessing the Harbison et al . ChIP-Chip dataset left 158 transcription factors and 2 , 779 target genes . Using a two step optimization algorithm developed by Liao et al . , we inferred the concentration profile for each transcription factor over the genetic gradient and compared it to the corresponding TF expression profile by computing Pearson's correlations ( ) . Figure S3 shows that these quantities were not well correlated with average correlation coefficients of and using the Lee et al . and Harbison et al . datasets respectively . The stability of the inferred TF concentrations were however robust when we compared results from the two ChIP-Chip datasets with a correlation coefficient of ( Figure S4 ) . The robustness was also verified by bootstrapping experiments [31] ( Results not shown ) . We next applied our method to study the mechanisms by which regulatory hotspots , genomic locations in yeast shown to be linked to the expression of many genes , perturb the underlying transcription regulatory networks . Although several transcription factors have been known to act as master regulators in yeast , it has been surprisingly shown in previous eQTL studies that only a few regulatory hotspots are located close to transcription factors . We hypothesized that although complex regulatory mechanisms upstream of transcription regulation such as signaling pathways exist , transcription factors ultimately mediate the global regulation of gene expressions . Using our framework , we tested our hypothesis by determining whether a regulatory hotspot is linked to the concentrations or promoter affinities of active transcription factors to achieving this regulation . To identify the regulatory hotspots , we performed simple linkage analysis on only a subset of genes that were NCA compliant ( see Methods ) . Similar to previous reports , only a few hotspots were located cis to any known transcription factors [1] , [2] . For example , a hotspot located on chromosome 12 spanning basepairs 600 , 000 to 680 , 000 was cis to HAP1 while another hotspot located on chromosome 3 spanning basepairs 60 , 000 to 100 , 000 was cis to LEU3 . Several approaches [20] , [23] have identified additional putative causal regulators , many of which are not transcription factors , corresponding to these regulatory hotspots . We first considered SNPs located in regulatory hotspots that perturbed the concentrations of active transcription factors to cause global differential expression . Extending the NCA model to incorporate SNPs as perturbations did not require changing the optimization procedure . As shown in Figure 2C , we first decomposed the inferred transcription factor concentration matrix from applying the original NCA algorithm , , into two matrices and segregated by a SNP . Next , we identified those transcription factors whose concentrations were linked to the SNP using a simple t-test , an example is shown in bold in Figure 2C , and assessed the significance of the linkage by a permutation scheme ( see Methods ) . Using both the Harbison et al . and Lee et al . ChIP-Chip binding data , we found many transcription factors whose concentrations were linked to at least one SNP . Table 1 lists those linkages occurring at regulatory hotspots and the corresponding transcription factors . In addition to having a strong linkage , we also required the transcription factors in the table to have at least 6 ( Lee et al ) or 7 ( Harbison et al ) downstream targets whose expression levels were significantly linked to the regulatory hotspot . A number of transcription factors known to act as global regulators were identified . Of particular note , we found HAP1 to be the mediator of hotspot 6 located on chromosome 12 spanning basepairs 600 , 000 to 680 , 000 using the Harbison et al . dataset; and YAP1 and LEU3 to be mediators of hotspot 3 located on chromosome 3 spanning basepairs 60 , 000 to 100 , 000 . GCN4 was also identified as a mediator of this hotspot using the Lee et al . dataset but it was only marginally significant using the Harbison et al . dataset ( Result not shown ) . These results are concordant with previous findings [2] , [23] . In particular , LEU2 has been previously implicated to be linked to hotspot 3 where an engineered deletion of the gene occurs . Figure 4 are heatmaps showing the strong correlations between concentration levels of transcription factors , HAP1 and LEU3 respectively , and the expression levels of their downstream targets linked to the respective regulatory hotspots . We next examined hotspot 2 , a hotspot that has been previously identified by brem et al . to regulate budding and daughter cell separation through the causal regulator AMN1 [9] . We identified four transcription factors , ACE2 , MBP1 , SKN7 and SWI4 , whose active concentrations were significantly linked to hotspot 2 in both datasets . Five other transcription factors responsible for cell cycle transitions , ABF1 , FKH1 , OAF1 , RAP1 and SWI5 were also found to be significant in the Harbison et al . dataset . Some of these transcription factors are known to interact with each other and have similar profiles such as ACE2 and SWI5; and MBP1 , SKN7 and RAP1 . Figure 3A and Figure 3B are heatmaps showing the strong correlation between the concentrations of transcription factors ( ACE2 and SWI4 ) and the expression levels of their direct targets linked to the hotspot . Our results are consistent with previous findings that suggest ACE2 as a causal transcription factor mediating the global regulation of the mitotic-exit network ( MEN ) by AMN1 [23] even though ACE2's direct targets were not overrepresented for any GO biological processes or functional groups . This is probably because many downstream transcripts of the MEN were not considered in this analysis because there's no direct ChIP-Chip evidence of binding between these transcripts and ACE2 . Another interesting regulatory hotspot , occurring at chromosome 12 basepairs 1 , 040 , 000 to 1 , 060 , 000 , was found by Brem et al . to regulate subtelomerically encoded helicases through the causal regulator SIR3 . We found two transcription factors , GAT3 and YAP5 , whose concentrations were linked to this hotspot using the Harbison et al . data . YAP5 was also significant using the Lee et al . data . Figure 3D and Figure 3C show the strong correlations between GAT3 and YAP5 concentrations and the expression profiles of their targets . Unlike the previous example , the targets of YAP5 were enriched for helicases ( ) and consisted of many genes with unknown function as represented by a significant enrichment for the GO annotation of “biological process unknown” ( ) . These results suggest a potential novel mechanism for the regulation of subtelomerically encoded helicases mediated by YAP5 and GAT3 . We next considered SNPs located in regulatory hotspots that perturbed the promoter affinities of transcription factors to cause global differential expression . Modeling these perturbations required an extension to the NCA model . As shown in Figure 2D , in addition to decomposing the transcription factor concentration and gene expression matrices , we also decomposed the promoter affinities matrix , into and where the only difference between the two is the column corresponding to the global promoter affinities of the transcription factor of interest as shown in bold . We identified perturbed networks of genes and transcription factors by deriving a likelihood ratio statistic that compared the extended model to the original NCA model . Since the extended model included additional parameters , namely different promoter affinities between populations , we expected it to always fit the data better . Thus to assess significance , we used a permutation scheme that randomized the decomposition of individuals while preserved the topology of the bipartite graph ( see Methods ) . We revisited the regulatory hotspots discussed in the previous section . We speculated that transcription factors whose promoter affinities were perturbed by a regulatory hotspot must interact with other transcription factors whose concentrations were perturbed by the same hotspot to induce global differential expression of the targets . The intuition being if the in vivo concentrations of a transcription factor is relatively stable , then it could still regulate gene expression by differentially binding to other transcription factors to form a complex . A transcription factor's binding affinity for promoters is then in part determined by the concentrations of its partnering transcription factors . This is exactly what we observed in our results . For example , we found that hotspot 6 which was shown to be linked to the concentrations of HAP1 was also linked to the promoter affinities of HAP4 . HAP1 and HAP4 are known to interact in a complex to regulate global respiratory gene expression . Similarly , hotspot 8 was linked to the concentrations of DIG1 and the promoter affinities of STE12 . DIG1 has previously been shown to code for an inhibitor of STE12 , a transcription factor involved in pheromone induction and invasive growth [32]–[34] We next examined how two hotspots discussed in the previous section also perturbed promoter affinities of transcription factors . Figure 4 and Table 2 show that hotspot 2 was linked to the promoter affinities of ACE2 , SWI4 and UME6 . Hotspot 2 was also shown in the previous section to be linked to the concentrations of ACE2 and SWI4 but not UME6 , see Figure S2 for the expression profiles of the downstream targets of UME6 . Consistent with our speculation , UME6 has been shown to interact with SWI4 and SWI4 has been shown to interact with itself . Furthermore , we see that AMN1 is a target of ACE2 suggesting that the regulation of the mitotic-exit network might be feedback in nature . Figure 4 also shows a similar network consisting of the two transcription factors whose concentrations linked to hotspot 7 , GAT3 and YAP5 . Notice that while YAP5's promoter affinities were linked to the hotspot also ( thick edges ) , GAT3's were not ( thin edges ) . Consistent with previous results , YAP5 has been shown to interact with itself to modulate gene expression . These results suggest that in some transcription factors , particularly those that interact with themselves , both promoter affinities and concentrations of the transcription factor could be perturbed by a regulatory hotspot . On the other hand , some transcription factors might not have their concentrations perturbed by a hotspot but because of interactions with another transcription factor , has their promoter affinities perturbed giving rise to global differential expression of their targets . Previous eQTL analyses have shown that the most significant linkages occur cis to genes [1] , [2]and often located or in LD with SNPs located in the promoter regions of genes harboring transcription factor binding sites [35] . Our model allowed us to determined if differences in expression of a single gene could be attributed to cis genetic variations perturbing the local affinities of transcription factors on the promoter . There is a direct similarity between these perturbations and those that affect global promoter affinities . As shown in Figure 2E , SNP3 perturbs the local affinities of transcription factors for the promoter of G3 . We modeled this affect by decomposing the matrix into and where the only difference between the decomposed matrices was the row corresponding to G3 , as shown in bold . We used a likelihood ratio statistic to choose between two different models and assessed the significance based on permuting the genotypes of the individuals . Of the small subset of genes examined , 2294 from using the Lee et al . dataset and 2779 from using the Harbison et al . dataset , we found ≈45% of the transcripts ( 972/2294 Lee , 1315/2779 Harbison ) linked to at least one SNP at a FDR of with using a standard t-test . Out of these linkages , ≈30% were cis ( 257/972 Lee , 331/1315 Harbison ) . These proportions are consistent with what has been reported [10] . We postulated that many cis linked loci found by previous analyses and confirmed by our analysis are in LD with causal SNPs located in promoter regions . We further postulated that such a causal SNP corresponds to a variation in the primary sequence of a transcription factor binding site that affects the promoter affinity of a transcription factor or a complex of transcription factors . This model is consistent with the idea that a genetic variation at regulatory regions of the genome can give rise to observed subtle differences in gene expression across populations . We identified a total of 138 and 174 genes which have their local promoter affinities affected by a SNP with a FDR of . Figure 5A shows that there is high concordance between those genes with significant cis linkages and those whose promoter affinities were perturbed . We did not expect all cis linkages to perturb promoter affinities . There are potentially other regulatory machinery that operate on intronic 3′UTRs and 5′UTRs . Next we compared the perturbed genes found using the Lee et al . dataset versus those found using the Harbison et al . dataset ( Figure 5B ) . At a FDR of , 72 significant genes were shared between the datasets and 168 genes were not . We suspected that the different results obtained from these two datasets can be attributed to differences in network topology . The two binding datasets often reported genes with different sets of bound transcription factors and transcription factors with different sets of targets making the estimates of certain quantities inconsistent . Additional discrepancies arose from different sets of genes having been eliminated from each analysis due to the criteria placed on the network topology .
Although there is a growing wealth of literature identifying putative causal regulators in yeast and mouse using statistical approaches , some of which integrate different sources of information , it is not clear by what mechanism genetic variations perturb the underlying regulatory networks to give rise to global differential expression . We have presented an integrated framework based on network component analysis that directly models how genetic variations perturb the concentrations and promoter affinities of transcription factors to cause the differential expression of their targets . Such a model differs from current eQTL analyses in that a direct , testable mechanism of transcription regulation is specifically considered . Although these networks are limiting , both in terms of the amount of biology they explain as well as the dependence on experimental data for their inference , a substantial set of genes ( ≈1/3 ) was still considered . In our analysis , we show that many genes with cis linkages are likely to be regulated by transcription factors binding differentially to their promoter regions . We also show two representative examples of the complex mechanism of achieving global differential expression of a large number of transcripts , where the regulation of transcription factors involve two distinct processes and maybe feedback in nature . Our approach specifically uses one variation of the NCA algorithm to infer the concentrations and promoter affinities of transcription factors . The key aspect of our approach is that we treat genetic variations as perturbations to an underlying regulatory network whose structure is already known . In theory , any NCA like approach [36]–[38] where a network is inferred from known data such as ChIP-Chip experiments , protein-protein interaction experiments or literature can be extended to take into account genetic variation . There are also some natural extensions to the framework we have presented . First , one is not limited to considering only genetic variation as a perturbation . Other forms of perturbation such as media condition and disease pathogenesis can as well be applied in this approach to identify the corresponding effect on the networks . Second , our method considers the perturbation of only one SNP . Although several approaches have been proposed to investigate the statistical interaction of multiple SNPs on a phenotype [11] , [39] , it would be interesting to study the mechanistic interactions of multiple perturbations on a transcription regulatory network .
We used the expression measurements ( 6 , 216 transcripts ) and genotyping data ( 2 , 956 SNPs ) collected over 112 segregants of yeast derived from two parental strains BY4716 and RM11-1a originally described by Brem et al . The gene expression data is available at GEO ( http://www . ncbi . nlm . nih . gov/projects/geo/ ) with the accession number GSE1990 . ChIP-Chip data from two datasets [27] , [28] were used to generate two different transcription regulatory networks at a p-value cutoff of 0 . 001 . Consistency was checked in each case by comparing the networks generated from using a cutoff of 0 . 01 and 0 . 001 . We next checked for NCA compliance as outlined [31] . We were left with a sub-network of 2 , 294 transcripts and 100 transcription factors after processing the Lee et al . dataset and 2 , 779 transcripts and 158 transcription factors after processing the Harbison et al . dataset . We first performed a standard t-test to compare the population means between the segregated expression profiles of a single gene by a given SNP . We assessed the significance of our linkages by performing a permutation test as described [40] . We then identified regulatory hotspots by dividing the yeast genome into 493 20 kb bins and counted the number of significant trans linkages to unique gene expression levels each bin contained from the standard t-test . We found a total of 430 significant trans linkages using the Harbison et al . data and 290 using the Lee et al . data . Assuming a Poisson process where the rare event of a trans linkage occurs at a rate of 0 . 87 ( 430/493 Harbison ) and 0 . 60 ( 290/493 Lee ) , the probability of observing >7 linkages in the largest bin using the harbison_transcriptional_2004 data is and the probability of observing >6 linkages in the largest bin using the Lee et al . data is . Because of the differences in the set of genes used in the different datasets , we constructed a set of 11 hotspots shared between the two . NCA was originally developed to analyze time series based gene expression data but can be easily adapted to analyze whole genome expression data collected from different individuals in a population . In both cases , the goal is to infer the concentrations of active transcription factors and the promoter affinities from the expression levels of the target genes . This inference is made possible when the partial topology of the interaction network between transcription factors and target genes is determined from genome-wide location analysis that detects the binding of transcription factors to DNA promoter regions ( ChIP-Chip ) . Figure 1B shows an example of a bipartite graph where the expression levels of five genes are determined by the concentrations and promoter affinities of the three transcription factors . Formally , given a matrix of dimension where we have collected the expression levels of genes from individuals . Each column represents a separate microarray experiment that measures the expression levels of all genes in one individual . NCA approximates the relationship between the concentrations of active transcription factors and gene expression levels by a log-linear model of the type: ( 1 ) where is the gene expression level for gene in individual , is the concentration of transcription factor in individual and is the affinity of transcription factor for the promoter of gene . We can take the log of Equation 1 and transform it into a matrix representation: ( 2 ) Here , is a matrix of dimension representing the concentrations of the transcription factors in the individuals and is a matrix of dimension representing the affinities of the transcription factors for the promoters of the genes and is a matrix of dimension representing the residual . NCA analysis without incorporating genetic information seeks to iteratively find and that minimizes the quantity: ( 3 ) Finding the least squares estimates of and is equivalent to finding the maximum likelihood estimates under the assumption that the are independent identically-distributed ( iid ) vectors with Gaussian noise . In our model , a genetic variation induces global differential expression either by perturbing the concentrations of a transcription factor or the promoter affinities of a transcription factor on all of its targets . Figure 1C shows the former case where the promoter affinities of TF1 on all targets remain the same but the concentration of TF1 is elevated in the group of individuals with an A allele at SNP1 while it is attenuated in the group of individuals with the C allele at SNP1 . Figure 1D shows the latter case where the affinities of TF2 for the promoter region of its targets are different between two populations . Notice that in both cases , we do not make any assumptions about where the genetic variation occurs since several mechanisms can contribute to the transcription factor having different in vivo concentrations and promoter affinities . We can formally model perturbations to the promoter affinities by constructing two matrices , and that differ in the column corresponding to the transcription factor of interest . We can also model local changes to the promoter affinities of all transcription factors on a single gene such as shown in Figure 1E where one group of individuals has the A allele and another group has the T allele ( SNP3 ) in the binding site of the transcription factor complex . To model this change in the promoter affinities on one gene , we again construct two matrices and that differ in the row corresponding to the gene of interest .
|
One of the fundamental challenges in biology in the post-genomics era is understanding the complex regulatory mechanisms that govern how genes are turned on and off . In a single organism where the functions of individual genes in a population do not differ much , many of the differences between individuals including physical phenotypes , susceptibility to disease , and response to drugs can be attributed to how genes are regulated . Previous studies have largely focused on identifying regulator and target genes whose expressions are linked to genetic variations in a population . We present work that focuses on considering a specific set of regulators called transcription factors whose targets can be verified from experiments and whose interactions with those targets have been well studied and modeled . In this setting , we can begin to understand how genetic variations perturb the concentrations and promoter affinities of active transcription factors to induce differential expression of the targets . Understanding the effects of these perturbations is important to understanding the fundamental biology of gene regulation and can help us to design and assess therapeutics and treatments for complex diseases .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"computer",
"science/applications",
"computational",
"biology/systems",
"biology",
"genetics",
"and",
"genomics/gene",
"expression",
"computational",
"biology/transcriptional",
"regulation",
"genetics",
"and",
"genomics/bioinformatics"
] |
2009
|
Using Network Component Analysis to Dissect Regulatory Networks Mediated by Transcription Factors in Yeast
|
The Middle East is a culturally and politically diverse region at the gateway between Europe , Africa and Asia . Spatial dynamics of the fatal zoonotic disease rabies among countries of the Middle East and surrounding regions is poorly understood . An improved understanding of virus distribution is necessary to direct control methods . Previous studies have suggested regular trans-boundary movement , but have been unable to infer direction . Here we address these issues , by investigating the evolution of 183 rabies virus isolates collected from over 20 countries between 1972 and 2014 . We have undertaken a discrete phylogeographic analysis on a subset of 139 samples to infer where and when movements of rabies have occurred . We provide evidence for four genetically distinct clades with separate origins currently circulating in the Middle East and surrounding countries . Introductions of these viruses have been followed by regular and multidirectional trans-boundary movements in some parts of the region , but relative isolation in others . There is evidence for minimal regular incursion of rabies from Central and Eastern Asia . These data support current initiatives for regional collaboration that are essential for rabies elimination .
Rabies is a fatal encephalitis caused by viruses in the genus Lyssavirus [1 , 2] . Although the majority of lyssavirus species are associated with bats , rabies virus ( RABV ) has successfully adapted to terrestrial carnivores on multiple occasions [3 , 4] and causes an estimated 70 , 000 deaths each year [5] . The majority of rabies cases in humans are caused by the bite of infected domestic dogs ( Canis lupus familiaris ) , but rabies can persist in both domestic dog and wildlife reservoirs [6] . In addition to the morbidity and mortality burden , costs are incurred through the necessity for provision of post-exposure prophylaxis and surveillance in rabies endemic areas , leading to an annual global economic cost estimated at over 500 million dollars [7 , 8] . Concerted control efforts in many regions have demonstrated the feasibility of rabies elimination in carnivores [6 , 7 , 9] . These control efforts are dependent on local epidemiology of the disease which will vary from region to region , depending on differing ecological and sociological factors [10] . The Middle East is a politically diverse region with a rich cultural history , situated between Europe , Asia and Africa . This position , and the region’s political and cultural variety , have had implications for the control of trans-boundary diseases of animals such as Foot and Mouth Disease and zoonotic diseases such as Avian Influenza , Brucellosis and Middle East respiratory syndrome coronavirus ( MERS-CoV ) [11–14] . Economic restrictions , conflict and political instability can also affect surveillance for diseases , in addition to causing acute and unpredictable human or animal migration [15 , 16] . Recent reported annual incidences of human rabies in countries of the Middle East vary from 0 . 02 to 1 . 3 per million human population , with annual incidence of post exposure prophylaxis administration varying from 1700 to over 6000 per million [16 , 17] . These figures are heavily influenced by variation in surveillance and reporting in different countries , yet reflect an on-going burden of rabies in the region [16–18] . Although many countries in the Middle East collate and report human rabies cases , routine surveillance and reporting of animal rabies is less systematic , often relying on local awareness and resources with associated potential biases [16 , 17 , 19] . There are multiple historical reports of disease consistent with rabies throughout the history of civilisation in the Middle East ( reviewed by [19] and [20] ) . Reconstruction of viral evolution using , molecular phylogenetics , suggests that the spread of RABV through human mediated dispersal from Europe is a dominant factor in the current epidemiology of rabies in the Middle East [4 , 21 , 22] . Dating this spread of a widely distributed lineage of RABVs , known as the ‘cosmopolitan’ lineage , using molecular clock analyses concurs with this scenario . The estimated date of divergence , 200–300 years ago , coincides with a period of increased human movements , increasing human population , and increasing domestic dog population associated with urbanisation in the 18th century . Current epidemiological evidence also supports this , suggesting dog rabies predominates , at least in Jordan , Lebanon , Iraq and Iran , [16 , 17 , 19 , 23–27] and also remains a concern in some central Asian countries and in the Caucasus region [18 , 28] . However , the historical reports of disease consistent with rabies in the Middle East pre-date this relatively recent spread of the cosmopolitan lineage , and additional separate rabies lineages appear to be spreading in neighbouring regions which may also be circulating in the Middle East [16 , 18 , 29] . Throughout the Indian subcontinent and South East Asia , dogs are considered the principal reservoir of rabies , with a genetically distinct lineage of RABV predominating , indicating prolonged circulation without mixing from other regions [30–33] . In addition , more recently a lineage of RABVs related to those endemic in Arctic regions has emerged , and then become the dominant lineage in parts of South Asia [29 , 31 , 34 , 35] . Establishing whether rabies in the Middle East today is , for example , the result of ancient lineages that have persisted in the region , or the result of incursion from other areas is essential in targeting sustainable control methods . Endemic dog rabies had been almost eliminated from Western Europe by the beginning of the 20th century , but in the second half of the 20th century an epidemic of fox ( Vulpes vulpes ) rabies spread through Europe , attributed to spill-over from dogs , most probably from more than one location in Eastern Europe [28 , 36–38] . With six wild canid species occupying the varied habitats in Western Russia and Eastern Europe , the epidemiology is complex [28] . The highest numbers of reported wildlife cases in Eastern Europe are in European red foxes and raccoon dogs ( Nyctereutes procyonoides ) . In the Middle East rabies has been reported in foxes , but also in golden jackal ( Canis aureus ) and wolf ( Canis lupus lupus ) [17 , 39] . Control methods for dog rabies and wildlife rabies vary considerably in required knowledge base , stakeholder involvement and cost [9 , 10] . Therefore a better understanding of the epidemiology is a fundamental prerequisite to planning rabies control . Previous phylogenetic studies of viruses detected in the Middle East have demonstrated probable trans-boundary movement of rabies , but have been unable to infer direction of movement and have been limited by numbers of samples available from some regions [16 , 22 , 27 , 36 , 40 , 41] . Here we have taken a broader and more comprehensive perspective on circulating RABV lineages in this particular region of the world by collecting RABV isolates from as many Middle East countries as possible as well as neighbouring regions . Through international collaborative efforts , we have obtained or generated sequence data for a panel of 183 RABVs spanning 40 years and from over 20 countries including previously un-sampled areas . We have also applied discrete phylogeographic techniques to infer migration events giving insight into rabies epidemiology in the region , with implications for control .
No animals were used for this study . Brain samples were analysed that had previously been taken for diagnostic purposes , from animals that were already dead , with the permission of the local competent authority . A panel of 183 RABVs were selected for analysis from over 20 countries spanning the years between 1972 and 2014 ( S1 Table ) . Sixty three virus sequences were derived in this study , with an additional 120 sequences derived in previous studies obtained from GenBank . Samples were collected as part of routine rabies sampling . This sampling will include rare systematic surveillance systems , but the majority of sequences will be from a laboratory-confirmed subset of ad-hoc reports of disease suspicion in either the public or animal health professionals . Samples include domestic dogs ( ownerless , owned and free roaming , and owned and restricted ) and wildlife . Virus sequences were derived from original clinical samples either in country of origin or in international reference laboratories ( FLI Germany , APHA UK and VetmedUni , Vienna , Austria ) on behalf of , and with permission of , local veterinary and public health authorities . Virus RNA was extracted from brain samples of rabid animals using TRIzol reagent , or using a guanidine based extraction kit ( RNeasy , Qiagen ) according to the manufacturer’s instructions . RNA was quantified by spectrophotometer and a 582 base pair region of the nucleoprotein ( N ) gene amplified by hemi-nested reverse transcriptase polymerase chain reaction ( hn-RTPCR ) as described previously ( APHA and FLI ) [42] or a combination of smaller RT-PCRs were used to generate a contiguous 948 base pair region of the nucleoprotein ( Vienna , primer sequences are available on request ) . PCR products were purified ( QiaQuick PCR Purification kit , Qiagen ) and sequenced using chain-termination ( Sanger ) sequencing ( Big Dye Sequencing kit , ABI ) . At least one forward and one reverse primer were used to generate a consensus sequence for each virus , which were visually checked for errors prior to alignment using the DNAStar package ( Lasergene ) . Sequences generated in this study were combined with selected published RABV sequences to include , as far as possible , comprehensive cover of the Middle East , and representative sequences from surrounding regions . Sequences were aligned using ClustalX2 ( version 1 . 2 ) . The length of sequence available for analysis varied among isolates , but a 400 base pair region of the nucleoprotein gene was chosen for further analyses as it was common to 171 of the selected sequences . A phylogenetic analysis of all these 171 virus sequences was implemented using Bayesian Markov Chain Monte Carlo simulation in the BEAST packagev1 . 8 . 0 [43] . A TN93 nucleotide substitution model with rate variation and a proportion of invariant sites ( Gamma+I ) were determined to best fit the data using Akaike Information Criterion in MEGA 6 . 0 . A relaxed and strict molecular clock model with either constant or Bayesian skyline population prior were used in Markov Chain Monte Carlo ( MCMC ) simulation for 100 , 000 , 000 iterations , sampling every 10 , 000 states to give effective sample sizes of over 200 . Molecular clock and population coalescent models were compared using a modified Akaike information criterion ( AICM ) in Tracer v1 . 5 on a subset of 137 sequences , as described previously [44] ( Table 1 ) , and maximum clade credibility trees were annotated using TreeAnnotator ( v1 . 8 . 0 ) after 10% of trees were discarded . The maximum clade credibility ( MCC ) trees were then visualised using Fig Tree ( v1 . 4 . 0 ) ( Fig . 1 ) . A strict molecular clock and Bayesian skyline population prior gave the lowest AICM and were therefore used for the estimates of divergence dates . Neighbour-joining and maximum likelihood phylogenetic reconstructions were also undertaken on the data set in MEGA6 [45] for comparison with Bayesian reconstructions . A neighbour joining tree was derived from p-distances with 1000 bootstrap replicates and a maximum likelihood tree using the same TN93 +G +I evolutionary model as used in the Bayesian analyses , the Nearest-neighbour inference heuristic and 100 bootstrap replicates . NJ and ML trees were visualized in FigTree ( v1 . 4 . 0 ) . An additional Bayesian MCMC phylogenetic analysis , with the same model parameters as above , was undertaken on a set of 59 sequences of 380 base pairs in length . These included 12 sequences from Oman and the United Arab Emirates which had only a 380 base pair region common to the other sequences , and were therefore not represented in the larger 400 base pair data set . These 12 were combined with 47 related sequences from the larger data set that were trimmed to 380 base pairs . A discrete phylogeographic analysis was also undertaken on a subset of 139 samples ( only viruses representing the ‘Cosmopolitan’ lineage ) . Samples were allocated to one of eight locations based on the country of isolation ( Europe , Caucasus , Middle East , Arabian Peninsula , Iran , Central Asia , Africa , Turkey ) ( Table 2 ) . These location groups were chosen to give approximately equal numbers of sequences per region and were based on geographical proximity of the recorded countries of origin . A data partition was generated from the location allocated to each sequence , meaning that each sequence is therefore identified by nucleotide sequence , date of isolation and location of origin . The same nucleotide substitution ( TN93+G+I ) , strict molecular clock and Bayesian skyline population coalescent models were used . The probable locations of each ancestral node were then reconstructed in the ensuing Bayesian MCMC analysis and displayed on the maximum clade credibility tree annotated using TreeAnotator ( v1 . 8 . 0 ) and visualised using FigTree ( v1 . 4 . 0 ) ( Fig . 2 ) . The strength of association between phylogenetic relationships and location of virus origin was assessed using an association index ( AI ) [46] . A null data set was created by randomisation of phylogeny and location data and an index of association between phylogeny and location in the observed data was compared to the equivalent from the null data set . The posterior set of trees produced by the Bayesian simulation was analysed with BaTS software ( after burn-in was removed ) , to incorporate uncertainty in calculation of the AI [47] . Ratios between the AI for the data under test , and AI for the null data set give an indication of the strength of association between genetic and spatial information . Ratios approaching one suggest that genetically related viruses are spatially homogenous , whereas low AI ratios imply a strong association between location and phylogenetic relationships . The number of trans-boundary movements or migration events suggested by the data were estimated using Markov jump counts [48] , and the support for those migrations were assessed using a Bayesian stochastic search variable selection procedure ( BSSVS ) , visualised using SPREAD [43] , with Bayes factors greater than 90 taken as strong support . A separate analysis was undertaken to infer the most probable reservoir for the ancestors of selected clades . In a similar manner to the phylogeographic analysis , each sequence was given a discrete trait corresponding to the species in which the virus sequences were detected . Due to the inherent uncertainty in whether the host a virus was detected in is the actual reservoir , all likely ‘dead-end’ hosts were removed . This stringent approach was taken to increase the reliability of the analyses at the expense of sample size , by only using sequences from domestic dogs ( including ownerless and free-roaming dogs ) ( n = 38 ) , and wild canids ( n = 35 ) . A TN93 G+I model of nucleotide substitution , a strict molecular clock and constant population size were used as model priors . The analysis was run for an MCMC chain length of 20 , 000 , 000 iterations and the chosen maximum clade credibility tree was annotated with TreeAnnotator ( v1 . 8 . 0 ) , visualised with FigTree ( v1 . 4 . 0 ) and coloured by host species with the highest posterior probability ( Fig . 3 ) . The approximate distribution of each detected RABV lineage was visualised using ArcGis ( ESRI ArcGis version 10 . 0 ) at the highest spatial resolution available for virus locations ( Fig . 4 ) .
These data support a clear and deeply rooted ancestral division of the represented RABVs into the well described ‘Cosmopolitan lineage’ , and the previously characterised ‘Asian dog’ and ‘Arctic/Arctic-like’ lineages ( Fig . 1 ) . The division into these clades is conserved between trees inferred using NJ , ML and Bayesian methods ( Figs . 1 , S1 , S2 ) . The strength of association between spatial and genetic relationships was assessed using the Association index . When compared to the null data set , the sample of RABVs showed a strong association between geographical origin and phylogenetic relationships ( AI ratio 2 . 11/11 . 52 ) , p<0 . 01 . Inference of non-negligible rates of migration through Bayesian stochastic search variable selection procedure ( BSSVS ) detected seven events within the study region , with Bayes Factor support over 90 ( Table 3 ) . These include migration between the Caucasus and other regions ( Iran , Europe and Central Asia ) , and migration between Turkey and all neighbouring regions ( Caucasus , Arabian Peninsula , Europe ) . Bayes Factor support for these migrations does not imply directionality , as a symmetrical network was assumed in the BSSVS procedure . However , combined with the discrete phylogeographic analysis these strongly support the picture of spread of rabies from Europe , Iran and Turkey into the other regions of the Middle East .
These analyses of a large number of RABV sequences from Europe , the Middle East and Asia provide a comprehensive overview of circulating RABV strains in domestic animals and wildlife , and show the geographic and temporal scale of trans-boundary movements of rabies within and between these regions . Previous studies have demonstrated likely spread in the region , but here we have been able to infer the direction of trans-boundary movements and make informed estimates of their frequency . These inferences have direct implications for rabies control policies and demonstrate that due to variation in patterns of rabies maintenance and spread a regional approach should be taken to reduce and eventually eliminate the disease . The first written record of disease consistent with rabies associated with dogs comes from Mesopotamia , a region corresponding to modern day Iraq [16 , 20] . Rabies continues to be a significant public and animal health issue in the region almost 4000 years later . We found no evidence that currently circulating strains were direct descendants of virus ancestors that occurred in the region 4000 years ago . In contrast , these analyses support at least one introduction of rabies from Europe with subsequent spread , albeit on a markedly different timescale , in the last 150 years . Indeed , estimates of the origin of all RABVs currently circulating in non-flying mammals are as recent as 1000 years ago . Our estimates of viral evolutionary rates in the partial nucleoprotein gene sequence concur with these previous analyses [4 , 32 , 36 , 50] . It is possible that ancestors of these ancient viruses are still circulating undetected , though a much more parsimonious explanation is that these ancient RABVs have been replaced by contemporary viruses [4 , 51] . A further consideration is that there is likely to be variability in selection pressures over time that nucleotide substitution models are unable to properly account for . This variability can lead to gross underestimates of the origins of viral clades which could explain the discrepancy between the historical records and molecular dating for RABV origins [52] . There is a clear and strongly supported distinction between those viruses circulating in Pakistan and Afghanistan ( Arctic-Like 1b ) , and those further west . Inclusion of recent isolates from Afghanistan add to previous studies [29 , 31 , 53] demonstrating that the dominant lineage in Pakistan and Afghanistan is closely related to Arctic RABVs . With the exception of two viruses detected in Iran also from the Arctic-like lineage , there is an apparent barrier to spread of the Arctic-like lineage at approximately 60 degrees longitude , corresponding to the Iranian border . Similarly , the Cosmopolitan lineage viruses appear to have spread widely to the west of 60 degrees longitude but not to the East , and it is currently unclear why . There is evidence that the RABV within the ‘Cosmopolitan’ lineage that are closely related to Chinese street strains are due to poorly attenuated vaccines [54] . Natural barriers such as high mountain ranges are unlikely to serve as a complete explanation here considering the fact that Arctic-like lineages appear able to cross the Himalaya [29 , 53] . The east of Iran is much less densely populated than the west , with large tracts of uninhabited land . With good evidence for human mediated dispersal of dog rabies in other regions [48] , human movements are likely to play a role in the Middle East . The nomadic pastoralist communities in parts of Iran and Central Asia that keep ‘choupan’ watch dogs [55] would be a potential route of trans-boundary movement of rabies , and also wildlife such as jackal ( Canis aureus ) and wolf ( Canis lupus ) are able to move freely across borders . Whether the Arctic-like lineage now dominant in Afghanistan , Pakistan and India continues to spread remains to be seen , but this apparent barrier to recent rabies spread between Iran and Afghanistan/Pakistan could be extremely important and deserves further investigation . In contrast with the apparent barrier to eastward spread at 60 degrees longitude , we have demonstrated significant support for movement of rabies among countries to the west of Iran . Notably , at least one of the clades of recent or currently circulating viruses sampled in this study ( Clade D ) had an ancestor that most probably occurred in Iran . Dispersal of dog rabies has been strongly associated with human movements [48 , 56] . The history of human migration in the Middle East is complex , dynamic and has been affected by civil unrest and conflict leading to poor reporting or recording of human movements [57 , 58] . In addition , comparatively little quantitative data are available for dog demographics . These two knowledge gaps limit understanding spread and therefore planning control strategies for dog rabies [59 , 60] . However , considering that the overwhelming majority of human cases are caused by bites from domestic dogs , systematic regular dog vaccination campaigns are currently still required to prevent human cases . Wildlife rabies presents a complex challenge , as demonstrated by the expensive and prolonged reduction of fox rabies in Western Europe through oral vaccination with live attenuated rabies vaccines [9] . In addition , as we have demonstrated here , wildlife and dog rabies are not as distinct as could be expected from this Western European experience [9 , 61 , 62] . For example , lineage B is wildlife-associated , but descendants of that lineage are now in dogs in Turkey and Azerbaijan . It is not possible to distinguish reliably with these data whether the samples are the result of repeated spill over from a wildlife reservoir or whether they represent maintenance of rabies in dogs . This is a critical question in rabies control , as demonstrated in parts of Africa where elimination of rabies in domestic dogs has led to the reduction of rabies in wildlife [63] . Both wildlife and dog rabies continue to be reported in Azerbaijan and Turkey , and epidemiological studies in Turkey and Israel strongly support spill over from wildlife [27 , 64–67] . Although control of dogs remains a priority from a public health perspective , eliminating rabies will require control in all reservoirs . Fortunately , with the advent of rapid and affordable sequencing [68] , and increased diagnostic testing and surveillance capacity in resource limited countries , it should be possible to resolve this important issue and establish whether there is independent maintenance of rabies in wildlife in these countries [69] . Historical records of rabies in the Middle East have often involved infection in gray wolves ( Canis lupus ) including an incident of one wolf infecting over 100 people in Adalia , Turkey in 1850 , and another significant event with multiple human fatalities in Iran in 1954 [70 , 71] . Wolves are not considered a natural reservoir due to their social structure and demography [72] but are an important vector of rabies to humans . For example , wolf bites historically featured disproportionately as the cause of human rabies cases in Russia , responsible for up to 56% of human cases toward the end of the 19th century [73] . The large size and ferocity of rabid wolves may lead to a disproportionate number of people and domestic animals affected , but also increase the likelihood of the event being detected and reported . With evidence of recent interbreeding between wolves and shepherd dogs in the Caucasus , and now the evidence for a dog associated strain in a wolf in Syria , there is potential for more complex reservoir dynamics [74] . Again , continued improvement of systematic surveillance and international reporting will help resolve these questions [16–18 , 39] . As with all studies using routine sampling for rabies , these data are limited by sporadic and inconsistent sample selection . This ad-hoc nature of most animal rabies sampling predisposes to bias towards animals of economic value , or those most likely to come into contact with humans . We cannot rule out the possibility of failing to sample all currently circulating lineages , nor account for the effect on the phylogeographic analysis of lineages that have become extinct . For example , epidemiological data supports the elimination of dog rabies from Europe in the latter part of the 20th century and those viruses are not represented here . The number and host species of samples from each country must also not be over-interpreted due to possible bias from differences in sampling strategy and the relative accessibility of samples from domestic animals or wildlife . For the discrete phylogeographic and reservoir host analyses , data were deliberately allocated into groups ( regions , dog/wildlife ) to improve the power of the analyses , at the expense of resolution . The countries were grouped based on geographic location and to give approximately equal numbers of sequences per group . Limited location data were available for many of the sequences ( country level only ) making continuous phylogeography unsuitable here , but if further location data were available the details of rabies dispersal would lend themselves to quantitative phylogeographic analyses . Considering this fact the distribution of RABV lineages identified in this study as shown in Fig . 4 are an approximation based on limited sequence location data . Nevertheless , mapping of the lineages allowed a more comprehensive view on the presumed geographic distribution in the Middle East that would be obtained from phylogenetic trees alone ( Figs . 1 , 2 ) . Our analyses were limited by available sequence length , with many of the sequences from historical samples having only a partial nucleoprotein gene sequence available and no original sample available for further sequencing . Comparison of relationships to those inferred using other regions such as the G-L intergenic region used in previous studies was therefore not possible . While this may restrict resolution , the analyses produced phylogenetic trees with strong posterior support for key nodes allowing reliable interpretation of those nodes , and major differences in relationships would not be expected , based on results from previous studies [38 , 75] . Dogs are responsible for the majority of human rabies cases; something that has been demonstrated conclusively in other regions , and where investigated , also in the Middle East [16 , 17 , 19 , 23–27 , 60] . Therefore adequate vaccine coverage in owned dogs should remain a priority [76] . However , these data show regular host switching and repeated trans-boundary movement of rabies demonstrating that long term control will require a more detailed understanding of dog demographics and human movements , a high resolution assessment of the role of wildlife and cross-species transmission in the maintenance of rabies in the region and a regional approach to rabies control . Current international collaborative efforts in the region tasked with improving surveillance and international reporting will facilitate achieving these aims [17 , 18 , 39] .
|
Despite being one of the oldest recognised infectious diseases , rabies continues to cause thousands of preventable human deaths per year . As a zoonotic disease , control of infection in the reservoir has been proven the most efficient route to reduction of human cases . In some regions , the epidemiology is well understood , with either dogs or wildlife known to be the primary reservoir and with little or no movement from , or into other regions . This is not the case in the Middle East , where rabies is underreported in animals and humans , there is little laboratory confirmation of infection , and the extent of rabies spread from country to country is not known . Previous studies have demonstrated trans-boundary movement of rabies but have been limited by a low number of available samples from some countries , and the direction of spread has been difficult to estimate . Here we use rabies virus partial genome sequences of 183 viruses from over 20 countries , combined with geographical and temporal information , to reconstruct the evolution of rabies viruses circulating in the Middle East . The results reveal an apparent barrier to spread between some regions but regular movement between others . These analyses will support policy on rabies control by indicating the relative importance of local control and animal movement restrictions when allocating resources .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Complex Epidemiology of a Zoonotic Disease in a Culturally Diverse Region: Phylogeography of Rabies Virus in the Middle East
|
Expression of E7 proteins encoded by carcinogenic , high-risk human papillomaviruses ( HPVs ) triggers increased expression of the histone H3 lysine 27 demethylase KDM6A . KDM6A expression is necessary for survival of high-risk HPV E7 expressing cells , including several cervical cancer lines . Here we show that increased KDM6A in response to high-risk HPV E7 expression causes epigenetic de-repression of the cell cycle and DNA replication inhibitor p21CIP1 , and p21CIP1 expression is necessary for survival of high-risk HPV E7 expressing cells . The requirement for KDM6A and p21CIP1 expression for survival of high-risk HPV E7 expressing cells is based on p21CIP1’s ability to inhibit DNA replication through PCNA binding . We show that ectopic expression of cellular replication factors can rescue the loss of cell viability in response to p21CIP1 and KDM6A depletion . Moreover , we discovered that nucleoside supplementation will override the loss of cell viability in response to p21CIP1 depletion , suggesting that p21CIP1 depletion causes lethal replication stress . This model is further supported by increased double strand DNA breaks upon KDM6A or p21CIP1 depletion and DNA combing experiments that show aberrant re-replication upon KDM6A or p21CIP1 depletion in high-risk HPV E7 expressing cells . Therefore , KDM6A and p21CIP1 expression are essential to curb E7 induced replication stress to levels that do not markedly interfere with cell viability .
Human papillomaviruses ( HPVs ) are a group of small , double-stranded DNA viruses that infect the squamous epithelium . The more than 200 HPV types described to date can be divided into mucosal and cutaneous types based on their tissue tropism . The mucosal HPVs can be clinically designated “low-risk” or “high-risk” based on their propensity to cause lesions that can undergo malignant progression . High-risk HPV infections account for approximately 5% of all human cancers , most notably cervical carcinomas , the third most common cancer in women worldwide [1 , 2] . Other anogenital tract cancers , including anal , vulvar , vaginal , and penile cancers , as well as oropharyngeal cancers , are also frequently associated with high-risk HPV infections [3 , 4] . The currently available prophylactic vaccines have no therapeutic efficacy . In addition , HPV-associated cervical cancers arise years to decades after the initial infection and vaccination rates remain low in many countries; as such , it will be decades before the current vaccination efforts will have a measurable impact on the incidence of HPV-associated tumors [5] . The E6 and E7 proteins are the major drivers of HPV-associated cancers , and persistent E6 and E7 expression is necessary for the survival of these tumors . E6 and E7 encode small non-enzymatic proteins that drive cancer formation by functionally re-programming cellular signal transduction pathways . The best known cellular targets of high-risk mucosal HPV E6 and E7 proteins are the p53 and retinoblastoma ( pRB ) tumor suppressors , respectively . Notably , these tumor suppressor pathways are also rendered dysfunctional by mutation in almost all human solid tumors [6 , 7] . Amongst the additional cellular targets of the HPV E6 and E7 oncoproteins that have been identified are enzymes that modulate histone modifications [8–17] . Dynamic post-translational modifications of histone tails impact both the physical state and the transcriptional competence of chromatin and play a critical role in the regulation of a variety of cellular processes such as stem cell maintenance , cell fate determination and maintenance , cell cycle control , and epigenetic heritability of transcriptional programs [reviewed in 18 , 19] . We previously reported that the repressive trimethylation of lysine 27 on histone H3 ( H3K27me3 ) , which is critical for epigenetic silencing mediated by polycomb group ( PcG ) proteins [20 , 21] is dramatically reduced in HPV16 E7-expressing primary human keratinocytes and in HPV16-positive cervical lesions and cancers [15 , 17] . The H3K27me3 mark is written by the histone lysine methyltransferase KMT6 ( EZH2 ) subunit of polycomb repressive complex 2 ( reviewed in [22] ) and erased by the histone lysine demethylases KDM6A ( UTX ) and KDM6B ( JMJD3 ) [23–27] , which are expressed at higher levels in these cells [15 , 17] . Although KDM6A and KDM6B appear identical with regards to catalytic activities and histone substrate specificities , KDM6A and KDM6B have non-overlapping and non-redundant biological activities . KDM6B may have both tumor suppressive and oncogenic activities in different cancer types . The KDM6B gene is located at 5q31 , an area that is frequently lost in various malignancies , including myeloid leukemias . However , KDM6B is expressed at high levels in prostate cancer , and its expression is further increased in metastatic prostate cancer [28] . Similarly , KDM6A appears to also have both tumor suppressive and oncogenic activities . Inactivating somatic KDM6A mutations have been detected in multiple cancers , including medulloblastoma , multiple myeloma , esophageal carcinomas , renal cell carcinoma , bladder cancer , and prostate tumors [29–31] . In contrast , KDM6A is rarely mutated in breast tumors , and activates oncogenic gene expression programs that control proliferation and invasion [32 , 33] . In a study of over 800 cervical and head and neck tumors from The Cancer Genome Atlas ( TCGA ) , HPV-positive tumors were found to express higher levels of KDM6A [34] . Moreover , KDM6B , but not KDM6A , regulates RAS/RAF-mediated oncogene-induced senescence ( OIS ) , one of several cell-intrinsic tumor-suppressor responses that function to eliminate aberrantly proliferating , potentially premalignant cells [35 , 36] . OIS is a major barrier to malignant progression , and additional genetic or epigenetic alterations are needed for progression to invasive cancer [37] . HPV16 E7 expression causes increased expression of both KDM6A and KDM6B , and HPV16 E7 expressing cells are dependent on KDM6A and KDM6B expression for cell survival [14 , 15] . Our previous studies revealed that the p16INK4A tumor suppressor is a critical downstream transcriptional target of KDM6B , and that p16INK4A expression is necessary for viability of high-risk HPV expressing cells [14] . Here we report that KDM6A expression is stimulated by high-risk , but not low-risk , HPV E7 proteins and that KDM6A expression is necessary for viability of high-risk , but not low-risk , HPV expressing or normal cells . We show that KDM6A controls expression of the cell cycle and replication inhibitor , p21CIP1 , and that KDM6A mediated induction of p21CIP1 expression is necessary for viability of high-risk HPV E7 expressing cells . We find that the ability of p21CIP1 to bind PCNA and inhibit DNA replication , rather than CDK2 binding and inhibition , is critical . Overall , our results suggest a model whereby p21CIP1 expression is necessary for the viability of HPV16 E7 expressing cells by dampening E7-induced replication stress .
It has been observed that KDM6A expression is increased in HPV16 E7 expressing cells [15 , 17] . To determine whether the ability to increase KDM6A expression was shared with other high-risk or low-risk HPV E7 proteins , we compared KDM6A levels in high-risk HPV16 E7 and HPV18 E7 expressing HFKs to HFKs expressing low-risk HPV6 or HPV11 E7 . HPV18 E7 expressing HFKs expressed KDM6A at similar levels as HPV16 E7 expressing cells , whereas low-risk HPV6 and HPV11 E7 expressing cells expressed KDM6A at levels similar to control vector transduced cells ( Fig 1A ) . The mRNA levels of the low-risk HPV E7s were higher than high-risk HPV E7s ( Fig 1B ) , and hence the ability of HPV E7 proteins is not related to expression levels but is a unique activity of the high-risk HPV E7 proteins . We next assessed KDM6A levels in several cervical cancer cell lines . KDM6A levels were higher than in primary human foreskin keratinocytes in the HPV16-positive SiHa and CaSki , the HPV18-positive HeLa , and the HPV39-positive Me-180 cervical cancer cell lines ( Fig 1C ) . Thus , KDM6A levels are generally higher in cervical cancer cells and cells expressing high-risk HPV E7 than in normal epithelial cells or cells expressing low-risk HPV E7 proteins . We have previously reported that depletion of KDM6A in the HPV16-positive cervical cancer line CaSki markedly inhibited cell viability [15] . To determine whether other cervical carcinoma cells showed a similar phenotype , we depleted KDM6A in the HPV16 positive cervical cancer lines SiHa and CaSki , the HPV39 positive cervical cancer cell line Me-180 , and the HPV18 positive cervical cancer cell line HeLa by transfection of a KDM6A-specific shRNA . Three different KDM6A shRNA expression vectors were used initially in SiHa cells ( Fig 2A ) , and shRNA 60 was chosen for the future experiments . KDM6A depletion was verified by qRT-PCR ( Fig 2B ) , and cell viability was assayed three days post-transfection . Consistent with our previously published results [15] , KDM6A depletion significantly decreases viability of SiHa cells , ranging from 36%; P = 0 . 0009 to 60%; P < 0 . 0001 . Similarly , KDM6A depletion in CaSki and Me-180 cells also caused significant 44% ( P = 0 . 0071 ) and 59% ( P < 0 . 0001 ) decreases in cell viability , respectively . In contrast , KDM6A depletion in HeLa cells caused a minor but non-significant 10% ( P = 0 . 1297 ) increase in cell viability , much like what we previously observed upon KDM6B depletion in this cell line [14] . ( Fig 2A ) . These results indicate that KDM6A expression is necessary for viability of some cervical carcinoma lines . The critical dependence of the viability of cancer cells on specific signaling pathways is generally referred to as “oncogene addiction” or “pathway addiction” [38] . By this definition , cervical carcinoma cells are addicted to KDM6A expression . The cervical carcinoma lines tested above contain integrated HPV genomes and consistently express the E6 and E7 oncoproteins . Given that HPV16 E7 causes increased KDM6A expression , we next determined whether HPV E7 oncoprotein expression was sufficient to generate KDM6A addiction . To test this hypothesis , we engineered donor- and passage-matched primary human foreskin keratinocyte ( HFK ) populations with ectopic expression of high-risk HPV16 E6 and/or E7 or HPV18 E7 and low-risk HPV6 or HPV11 E7 using retroviral vectors and verified expression by qRT-PCR and Western blot ( Fig 2D and 2E ) . KDM6A was depleted by infection with a lentiviral shRNA expression vector , depletion was verified by qRT-PCR ( Fig 2F ) , and cell viability was assessed . We observed a significant 43% ( P = 0 . 0052 ) decrease in cell viability in HPV16 E7-expressing HFKs . Similarly , cell viability was also significantly decreased by 64% ( P <0 . 0001 ) in cells that , like cervical carcinoma cells , co-express HPV16 E6 and E7 . In contrast , HPV16 E6-expressing HFKs and control-vector–infected HFK populations were not significantly affected ( P = 0 . 6767 and P = 0 . 1824 , respectively ) by KDM6A depletion ( Fig 2C ) . KDM6A depletion in HPV18 E7 expressing HFKs caused a significant 65% ( P = 0 . 0002 ) decrease in cell viability , while HPV6 and HPV11 E7 expressing HFKs were not significantly inhibited ( P = 0 . 9206 and 0 . 9863 , respectively ) ( Fig 2G , 2H and 2I ) . In addition to cell viability assays , we also determined cell numbers by Sulforhodamine B ( SRB ) assays following infection of SiHa cervical carcinoma cells with the recombinant KDM6A shRNA 60 expressing lentivirus at ten days after puromycin selection ( Fig 2L ) . These experiments revealed that cell numbers were significantly decreased by 65 . 4% ( P <0 . 0001 ) . Moreover , FACS analysis of SiHa cervical cancer cells showed a significant 17 . 6% ( P = 0 . 0003 ) increase in cells with a sub G0/G1 DNA content , supporting the notion that KDM6A depletion in SiHa cells causes apoptotic cell death ( Fig 2M ) . In summary , these results show that high-risk HPV E7 expression is sufficient to cause KDM6A addiction . In contrast , low-risk HPV E7 expression , which does not trigger KDM6A expression , does not generate KDM6A addiction . Since E6 does not markedly modulate sensitivity to KDM6A depletion , and given that E6 and E7 are the only HPV proteins consistently expressed in cervical carcinoma lines , these results show that KDM6A addiction of cervical carcinoma cells arises as a consequence of E7 expression . To determine whether KDM6A addiction is generated as an immediate consequence of HPV16 E7 expression or whether it is acquired after long-term E7 expression , we performed KDM6A depletion experiments in U2OS osteosarcoma cells with doxycycline-inducible HPV16 E7 expression . We showed previously that these cells express HPV16 E7 and KDM6A with a concomitant decrease of the H3K27me3 mark within 48 to 72 hours of doxycycline treatment . This is reversed when doxycycline is removed [15] . Depletion of KDM6A with either shRNA or siRNA duplexes did not significantly inhibit the viability of these cells before HPV16 E7 induction ( P = 0 . 1455 and 0 . 053 , respectively ) . In contrast , KDM6A depletion with shRNA or siRNA duplexes caused a significant 24% ( P = 0 . 0043 and 0 . 0003 , respectively ) decrease in viability after HPV16 E7 expression was induced by 72 h of doxycycline treatment ( Fig 2J ) . Depletion of KDM6A was verified by immunoblot ( Fig 2K ) . This result shows that KDM6A addiction arises as a direct and immediate consequence of HPV16 E7 expression . This finding is particularly remarkable because , unlike HPV-expressing cervical cancer cells , HPV E7 expression is not necessary for survival or even proliferation of U2OS cells . High-risk HPV E7 also induces expression of the related histone demethylase KDM6B and triggers KDM6B addiction [14 , 15] . Since both of these enzymes erase H3K27me3 marks , we determined whether KDM6A and KDM6B addiction was mediated through the same downstream pathway and/or if it involved similar mediators . To address this issue , we first investigated whether ectopic KDM6B expression may rescue the loss of viability caused by KDM6A depletion in CaSki cervical cancer cells . Ectopic expression of KDM6B , which was verified by qRT-PCR ( Fig 3B ) , did not inhibit the loss of viability upon KDM6A depletion ( P = 0 . 1968 , SiHa ) , while ectopic expression of a non-targetable KDM6A rescued the effects of KDM6A depletion ( Fig 3A and 3B ) . Similarly , ectopic expression of the downstream effector of KDM6B-addiction , p16INK4A ( CDKN2A ) , in these cells did not rescue the loss of cell viability caused by KDM6A depletion ( P = 0 . 1415 ) ( Fig 3A and 3B ) . CDK4/6 depletion also did not override the effect of KDM6A depletion ( P = 0 . 325 ) ( Fig 3C and 3D ) . These results show that KDM6A addiction is mediated by different downstream targets than those that mediate KDM6B addiction and specifically that KDM6A addiction is not related to p16INK4A and CDK4 and/or CDK6 inhibition . Despite the fact that p16INK4A and related CDK4/CDK6 inhibitors do not mediate KDM6A addiction , KDM6A depletion in high-risk HPV E7 expressing cells caused a similarly strong loss of cell viability as KDM6B depletion . Hence , we assessed whether cell cycle inhibitors other than p16INK4A may be involved . It has been reported that KDM6A can modulate expression of the CDK2 inhibitor , p21CIP1 ( CDKN1A ) [39] . Interestingly , p21CIP1 levels are high in HPV E7 expressing cells , and E7 has been reported to dampen its inhibitory activities [40 , 41] . If p21CIP1 was key to KDM6A addiction , HPV E7 expressing cells would also be addicted to p21CIP1 expression . To test this hypothesis , we depleted p21CIP1 in the HPV16 positive SiHa and CaSki and the HPV39 positive Me-180 cervical carcinoma lines , and assessed cell viability . Depletion of p21CIP1 depletion with multiple different shRNAs was verified by immunoblotting and qRT-PCR ( Fig 4B and 4C ) and significantly decreased viability of CaSki ( ranging from 66%; P = 0 . 0003 to 84%; P < 0 . 0001 ) , SiHa ( ranging from 81%; P < 0 . 0001 to 96%; P < 0 . 0001 ) , and Me-180 cells ( ranging from 83%; P < 0 . 0001 to 86%; P < 0 . 0001 ( Fig 4A ) . These results revealed that HPV positive cervical cancer cells are addicted to p21CIP1 expression and suggest that p21CIP1 mediates KDM6A addiction . Since high-risk HPV E7 oncoprotein expression was sufficient to generate KDM6A addiction , we hypothesized that high-risk HPV E7 expression may also be sufficient to generate p21CIP1 addiction . To test this hypothesis , we utilized primary HFK populations with ectopic expression of the high-risk HPV16 or HPV18 E7 or the low-risk HPV6 or HPV11 E7 proteins ( Fig 2D and 2H ) . p21CIP1 was depleted by infection with lentiviral shRNA expression vectors , depletion was verified by qRT-PCR ( Fig 4E ) , and cell viability was assessed . We observed significant 58 . 8% ( P = 0 . 0199 ) and 83 . 2% ( P <0 . 0001 ) decreases in viability upon p21CIP1 depletion in HPV16 E7 and HPV18 E7-expressing HFKs , respectively . In contrast , low-risk HPV6 and 11 E7 expressing HFKs were not significantly affected ( P = 0 . 4126 and 0 . 2908 , respectively ) ( Fig 4D ) . These results show that high-risk HPV E7 expression is sufficient to cause p21CIP1 addiction . To determine if p21CIP1 addiction arises as a direct and immediate consequence of HPV16 E7 expression , we depleted p21CIP1 in osteosarcoma cells with doxycycline-inducible expression of HPV16 E7 . Depletion of p21CIP1 ( Fig 4G ) did not significantly inhibit the viability of these cells before HPV16 E7 induction ( P = 0 . 1346 ) . In contrast , p21CIP1 depletion caused a significant 67% ( P < 0 . 0001 ) decrease in viability after HPV16 E7 expression was induced by 72 h of doxycycline treatment which was rescued by ectopic expression of a non-targetable p21CIP1 ( Fig 4F ) . Ectopic p21CIP1 expression rescued the decrease in viability of KDM6A depleted CaSki cervical cancer cells ( P < 0 . 0001 , KDM6A depletion compared to KDM6A depletion plus p21CIP1 expression ) and HPV16 E7 expressing HFKs ( P < 0 . 0001 , KDM6A depletion compared to KDM6A depletion plus p21CIP1 expression ) ( Fig 4J and 4K ) . In contrast however , ectopic p21CIP1 expression did not rescue inhibition of viability of KDM6B depleted CaSki cervical cancer cells ( P = 0 . 472 KDM6B depletion compared to KDM6B depletion plus p21CIP1 expression ) ( Fig 4L and 4M ) . Moreover , KDM6A depletion in CaSki and SiHa cells resulted in decreased p21CIP1 protein expression ( Fig 4N ) and caused a decrease of the repressive H3K27me3 mark at the CDKN1A promoter ( Fig 4O ) , functionally linking KDM6A and p21CIP1 . In addition to cell viability assays , we also determined cell numbers by SRB assays following infection of CaSki cervical carcinoma cells with the recombinant p21CIP1 shRNA 125 expressing lentivirus at ten days after puromycin selection ( Fig 4H ) . These experiments revealed that cell numbers were significantly decreased by 96 . 43% ( P = 0 . 02 ) . Moreover , FACS analysis of SiHa cervical cancer cells showed a significant 7 . 12% ( P = 0 . 0012 ) increase in cells with a sub G0/G1 DNA content , supporting the notion that KDM6A depletion in SiHa cells causes apoptotic cell death ( Fig 4I ) . In summary , these results show that high-risk HPV E7 expressing cells are addicted to p21CIP1 expression and that p21CIP1 is a major mediator of KDM6A addiction . The cyclin-dependent kinase ( CDK ) inhibitor p21CIP1 has amino-terminal cyclin and CDK-binding motifs [42–44] and a carboxyl-terminal PIP box that mediates binding to the proliferating cell nuclear antigen ( PCNA ) [45 , 46] . Despite the fact that both p21CIP1 is expressed at high levels in HPV-positive cells , HPV E7 interferes with the ability of p21CIP1 to inhibit CDK2 activity and PCNA-dependent DNA replication [40 , 41 , 47] . If p21CIP1 addiction of HPV E7 expressing cells were related to CDK2 inhibition , ectopic expression of p27KIP1 ( CDKN1B ) , which has similar CDK2 binding site and CDK2 inhibitory activity as p21CIP1 but does not inhibit replication through PCNA binding , would be predicted to rescue the loss cell viability caused by depletion of either KDM6A or p21CIP1 . However , this was not observed ( Fig 5A and 5B ) . Similarly , ectopic expression of a dominant negative CDK2 mutant [48] did not rescue the loss of cell viability caused by KDM6A depletion ( Fig 5C and 5D ) . Moreover , ectopic expression of the amino-terminal p21CIP1 fragment that contains the CDK2 binding domain and is sufficient for CDK2 inhibition [42] did not rescue the loss of cell viability in response to KDM6A depletion ( Fig 5C ) . Collectively , these results show that p21CIP1 addiction of high-risk HPV E7 expressing cells is not related to CDK2 inhibition . In contrast , ectopic expression of the carboxyl terminal p21CIP1 domain , which contains the PCNA binding site and is sufficient to inhibit PCNA [42] , efficiently rescued the decrease in viability upon KDM6A depletion . Moreover , expression of p21CIP1 carrying a mutation in the PCNA-interacting protein ( PIP ) box that disrupts PCNA binding and inhibition [49] did not rescue the decrease in cell viability upon KDM6A depletion ( Fig 5C ) . Collectively , these results show KDM6A and p21CIP1 addiction of high-risk HPV E7 expressing is related to PCNA binding and independent of CDK2 inhibition . We next attempted to directly assess the necessity of PCNA for p21CIP1 addiction . However , treatment of cells with T2 amino alcohol ( T2AA ) , a compound that disrupts interaction of PCNA with PIP box binding proteins , caused extensive cell death even in normal cells . This is not surprising , given that PCNA has multiple functions in DNA replication and DNA repair [50] . To nevertheless address whether the ability of p21CIP1 to inhibit replication is key to addiction of high-risk HPV E7 expressing cells to KDM6A , we assessed whether depleting the DNA licensing factors CDC7/DBF4 in SiHa cervical cancer cells might abrogate the KDM6A-mediated decrease in cell viability . The CDC7/DBF4 complex acts as a protein kinase that is required for the initiation of DNA replication [51 , 52] . Depletion of CDC7/DBF4 significantly inhibited the decrease of viability in response to loss of KDM6A expression . ( Fig 6A and 6B ) . Similarly , depletion of the pre-replication complex component , CDT1 , rescued the loss of viability caused by p21CIP1 depletion ( Fig 6C and 6D ) . These results support our hypothesis that the loss of viability of high-risk HPV expressing cells caused by KDM6A and p21CIP1 depletion are mechanistically connected to induction of aberrant cellular DNA replication . Aberrant firing of replication origins results in replication stress , a frequent hallmark of cancer cells . Replication stress triggers double strand DNA breaks and causes activation of the ATM and ATR kinases . These in turn causes genomic instability , and in severe cases when the DNA damage cannot be repaired , cell death . ATR is hyperactive in HPV E7 expressing cells [53–56] , and HPV16 E7 is known to induce replication stress due to aberrant E2F activity as a consequence of pRB inactivation and through other mechanisms [57] . Uncontrolled firing of replication origins can cause nucleotide starvation , and exogenous nucleoside supplementation has been shown to attenuate the replication stress response [58–60] . We hypothesized that high-level p21CIP1 expression in E7 expressing cells may dampen replication stress , presumably by complexing PCNA . Therefore , we tested whether dampening replication stress by nucleoside supplementation might rescue the loss of cell viability in response to p21CIP1 depletion . U2OS osteosarcoma cells with doxycycline-inducible HPV16 E7 expression were supplemented with 50 μM nucleosides in the media or grown under standard conditions . In contrast to cells grown under standard conditions , p21CIP1 depletion in U2OS osteosarcoma cells with doxycycline-inducible HPV16 E7 expression did not cause a significant loss of viability in nucleoside-supplemented cells ( Fig 6E and 6F ) . These results are consistent with a model whereby KDM6A and p21CIP1 addiction is based on p21CIP1 dampening replication stress of high-risk HPV E7 expressing cells . Replication stress causes single and double strand DNA breaks that are sensed by the ATR and ATM kinases . 53BP1 nuclear bodies are markers of DNA breaks , including those induced by replication stress . [61] . To determine whether KDM6A and/or p21CIP1 induction by E7 affected the incidence of DNA double strand breaks , we evaluated the number of 53BP1 nuclear bodies in KDM6A and p21CIP1 depleted SiHa cervical cancer cells . Depletion of KDM6A or p21CIP1 caused a 3 . 5 fold ( P = 0 . 0177 ) or 3 . 7 fold ( P = 0 . 0172 ) increase in 53BP1 nuclear bodies , respectively ( Fig 7 ) . These results further support the hypothesis that KDM6A and p21CIP1 addiction is based on the role of p21CIP1 in dampening replication stress in E7 expressing cells . To further analyze the hypothesis that KDM6A and p21CIP1 addiction is based on the role of p21CIP1 in dampening replication stress in E7 expressing cells , we analyzed replication in individual DNA fibers in KDM6A and p21CIP1 depleted SiHa cervical cancer cells . Depletion of KDM6A or p21CIP1 caused an increase in the amount of re-replication , as detected by yellow tracks ( average Pearson correlation coefficients = 0 . 16 ( control cells ) , 0 . 48 ( KDM6A-depleted cells ) , and 0 . 34 ( p21CIP1-depleted cells ) , adding further support to the hypothesis that KDM6A and p21CIP1 addiction is based on the role of p21CIP1 in dampening replication stress in E7 expressing cells ( Fig 8 ) .
Similar to oncogenic mutations in cellular oncogenes and tumor suppressors , infections with oncogenic viruses trigger innate tumor suppressor pathways . The evolution of these viruses , however , has been driven by the need to overcome such cellular defense responses . Similar to what has been reported for the MYC oncogene [62–65] , high-risk HPV E7 oncogene expressing cells are predisposed to undergo cell death particularly under conditions of limited growth factor availability , but high-risk HPV E6 proteins that are co-expressed during a viral infection effectively abrogate this response [66 , 67] . Similarly , we have shown that high-risk HPV E7 induces p16INK4A expression , similar to what has been reported for RAS expression causes “Oncogene-Induced-Senescence” ( OIS ) which is mediated by epigenetic de-repression of p16INK4A through the H3K27me3-specific histone demethylase KDM6B [14 , 15 , 35 , 36] . To evade elimination by OIS , high-risk HPVs have evolved to target the key OIS mediator , RB1 , for proteasomal degradation [68–71] . Most remarkably , HPV E7 expressing cells , including some cervical carcinoma cell lines become acutely “addicted” to KDM6B and p16INK4A expression [14] . While “oncogene-addiction” is a well-known concept [38] , addiction of cancer cells to expression of a tumor suppressor , such as p16INK4A , appears counterintuitive , at best . HPV16 E7 expression also causes increased KDM6A expression and addiction to KDM6A . Like KDM6B , KDM6A erases H3K27me3 repressive marks , thereby counteracting polycomb repression [23 , 26 , 27] . Despite this similarity in enzymatic activities , the two enzymes are involved in de-repressing distinct cellular targets [23 , 24 , 26 , 27 , 72–76] . Here , we show that KDM6A addiction is mechanistically distinct from KDM6B and is mediated by the CDK2 and PCNA inhibitor , p21CIP1 . It is likely that , similar to what has been reported for KDM6B , increased KDM6A expression also represents a cellular defense response to expression of high-risk HPV E7 proteins . KDM6A mediated epigenetic de-repression of p21CIP1 will inhibit cell cycle progression through CDK2 inhibition and DNA replication through binding and inhibiting PCNA . High-risk HPVs , however , have evolved to short-circuit this growth-inhibitory cellular defense response by dampening the cell cycle and replication inhibitory activities of p21CIP1 as well as through other mechanisms including enhanced expression of E2F regulated genes , many of which stimulate cell cycle progression and DNA replication ( Fig 9 ) . The ability of E7 to abrogate the CDK2 inhibition by p21CIP1 may be key to retaining differentiating keratinocytes in a replication competent state [77 , 78] , which is an essential requirement for HPV genome amplification and progeny synthesis in differentiated keratinocytes . Our experiments , however , yielded no evidence that the CDK2 inhibitory activity of p21CIP1 was rate limiting for KDM6A addiction . Specifically , ectopic expression of p27KIP1 , which contains a structurally related CDK2 interacting/inhibitory domain , or a kinase defective , dominant negative CDK2 mutant did not rescue the antiproliferative effect of KDM6A depletion . Rather , we discovered that the ability of p21CIP1 to bind and inhibit PCNA was necessary . PCNA is an E2F-responsive gene that is highly expressed in HPV E7 expressing cells and forms a trimeric “sliding clamp” that enhances the interaction of the DNA polymerase with the template DNA , thereby enhancing the processivity of the replicative DNA polymerase and coordinating DNA damage repair during DNA replication ( reviewed in [79 , 80] ) . In addition , post-translationally modified versions of PCNA also play important roles in post-replicative DNA damage repair . The p21CIP1 protein binds to PCNA through a PIP box , a conserved motif that is present in various PCNA interacting proteins that are involved in DNA replication . E7 , in turn , has been reported to bind p21CIP1 thereby abrogating the PCNA inhibitory activity of p21CIP1[41] . Given that p21CIP1 is highly expressed in E7 expressing cells [81–83] , and that p21CIP1 inhibition by E7 is based on a stoichiometric interaction [40 , 41] , it is likely that there is a significant pool of “active” , PCNA inhibitory p21CIP1 in E7 expressing cells . Our findings that ( a ) the PIP box in p21CIP1 is necessary to rescue the loss of cell viability caused by KDM6A depletion and ( b ) that the loss of cell viability caused by KDM6A or p21CIP1 depletion is abrogated by simultaneous depletion of replication factors are consistent with the model that high-risk HPV E7 expressing cells are addicted to PCNA inhibition by p21CIP1 . Why , though , would it be essential to inhibit PCNA in high-risk HPV E7 expressing cells when the major challenge of the viral life cycle is to keep the cellular replication machinery active for viral genome amplification and progeny synthesis ? Expression of replication regulatory proteins is a highly choreographed process that is tightly linked to the cell division cycle . Expression of many of these factors , including PCNA , is modulated by members of the E2F family of transcription factors . HPV E7 is well known to subvert E2F regulation by binding and enhancing the degradation of the RB1 tumor suppressor and the related RBL1 ( p107 ) and RBL2 ( p130 ) proteins as well as through other mechanisms [16 , 68–71 , 84] . Therefore , E2F-regulated replication factors , including PCNA , are highly expressed in E7 expressing cells . Aberrant firing of replication origins can lead to a process referred to as “replication stress” , which arises when replication forks stall . Single stranded DNA within these structures will trigger activation of the ataxia telangiectasia and Rad3 related kinase ( ATR ) . ATR activation is critical to resolve replication stress and to limit re-replication by engaging an S-phase checkpoint that is mediated by the TP53 as well as the RB1 tumor suppressor pathways [85 , 86] . In cells that have suffered TP53 and/or RB1 mutations , replication stress can generate genomic instability . Replication stress is a frequent hallmark of human tumors , and HPV16 E7 expression in normal diploid human cells induces hallmarks of replication stress [57 , 59] and double strand DNA breaks [87] . Most remarkable , E7-induced genomic instability can be ameliorated by supplementation with exogenous nucleosides , suggesting that E7 induced double strand DNA breaks and the ensuing genomic instability is at least in part triggered by replication stress [88] . E7-induced replication stress in differentiated epithelial cells , however , may be beneficial to the viral life cycle since activation of the double strand DNA break response , including ATM and ATR , enhance viral genome maintenance and amplification [53–55 , 89–91] . Our results suggest a model whereby p21CIP1 and specifically its ability to inhibit PCNA is necessary to limit high-risk HPV E7 induced replication stress to levels that cells can tolerate and do not markedly interfere with viability . Depletion of p21CIP1 in E7 expressing cells , however , enhanced replication stress and the associated double strand DNA breaks as evidenced by 53BP1 nuclear foci and the resulting of cell viability . The fact that nucleoside supplementation , which alleviates E7-mediated replication stress , can effectively dampen the observed loss of viability in response to p21CIP1 depletion strongly supports this model . Our results also suggest a model whereby KDM6A expression by an oncogenic insult caused by E7 expression . The cell cycle and replication inhibitor p21CIP1 is a rate-limiting component of this KDM6A-mediated cellular defense response that removes a potentially pre-carcinogenic cell from the proliferative pool by shutting down S-phase entry and DNA replication . High-risk HPVs , however , have evolved to evade this response by dampening the CDK2 as well as PCNA inhibitory activities of p21CIP1 through multiple mechanisms . Consequently , despite high-level p21CIP1 expression , high-risk HPV E7 expressing cells retain CDK2 activity and remain proliferatively active , but this generates some replication stress . E7-triggered replication stress in differentiating epithelial cells is not only an acceptable price to pay for the resulting abundant availability of cellular replication factors , but HPVs have evolved to take advantage of the resulting ATR/ATM activation for the replication of their genomes [53 , 89] . Nonetheless , high-level p21CIP1 expression is essential for viability of high-risk HPV E7 expressing cells since it keeps replication stress at a manageable level; loss of KDM6A or p21CIP1 causes cell death that is at least in part due to replication stress . Since many cervical carcinoma lines remain KDM6A and p21CIP1 addicted , this suggests that this pathway may be targetable for therapeutic intervention . In summary , our results show that KDM6A induction represents a cellular defense response to HPV E7 oncogene expression that is mechanistically independent and different from the KDM6B mediated response that we discovered previously [14] . KDM6B triggers p16INK4A expression causing activation of the RB tumor suppressor pathway which signals cellular senescence , whereas KDM6A activates p21CIP1 expression which corresponds to the cell cycle and replication inhibitory arms of the p53 tumor suppressor pathway . In each case , the virus had to adapt and evolve strategies to overcome these abortive cellular responses . In the case of KDM6B , high-risk HPV E7 proteins cause RB1 destabilization , thereby short-circuiting the RB1-mediated senescence response . In the case of KDM6A , high-risk HPV E7 proteins inhibit the cell cycle and replication inhibitory activities of p21CIP1 . This causes replication stress and DNA breaks , which activates the DNA repair machinery that these viruses harness for their own genome replication . Dysregulated cell cycle entry as a consequence of RB1 loss and CDK2 hyperactivity causes hyperproliferation that , in concert with the increased incidence of double strand DNA breaks , can cause genomic instability and malignant progression . The HPV-mediated subversion of the tumor suppressive activities of KDM6A and KDM6B and the ensuing addiction to these two enzymes of high-risk HPV expressing cells , but not normal cells , is not only of academic interest , but also provides novel therapeutic targets for high-risk HPV-associated lesions and cancers .
Primary HFKs were isolated and cultured as previously described [16] . HFKs were transduced by recombinant retroviruses carrying either the control vector ( LXSN ) or vectors encoding HPV16 E6 , HPV16 E7 , HPV16E6 and E7 , or HPV18 E7 [92] . HPV16 and 18 E6 and E7 expression was assessed by quantitative RT-PCR as previously described [14] . U2OS-tet on ( Clontech ) , CaSki ( ATCC ) , SiHa ( ATCC ) , Me-180 ( ATCC ) , and HeLa cells ( ATCC ) were maintained as previously described [14] . For experiments on the inhibition of PCNA interaction , cells were treated with T2AA ( Sigma ) for 72 h . The human keratinocytes used in this study were obtained from discarded foreskin circumcisions from anonymous donors at Brigham and Women’s Hospital and are not classified as human subjects research . These specimens were not specifically collected for this study and lack all identifiers . Uridine ( Sigma ) and cytidine ( Sigma ) were dissolved in distilled water to make 10 mM stocks , adenosine ( Sigma ) and guanosine ( Sigma ) were dissolved to make 2 mM stocks , and the suspensions was briefly boiled , filter sterilized , and added to complete medium at a final concentration of 50 μM . Transient transfections were performed using Polyethylenimine ( PEI ) ( Polysciences ) as described [93] . The cells were transfected with shRNA constructs and plasmids described in S1 and S2 Tables . Three days after transfection , media was removed , and 10 μg/ml resazurin sodium salt ( Sigma; diluted in growth medium ) was added to each well . The plates were incubated for 1–3 h at 37°C and then read at 570 and 600 nm on a microtiter well plate reader ( Biotek ) . To assess longer term viability , cells were selected with 1 μg/mL of puromycin at 24 hours post infection . At 10 days post-selection , the surviving cells were stained with sulforhodamine B ( Sigma ) , and quantified in a plate reader [94] . Total RNA was extracted using the Quick-RNA MiniPrep Kit ( Zymo ) , and cDNA was reverse transcribed using Taqman® Reverse Transcription Reagents ( Life Technologies ) . Quantitative RT-PCR was performed using either Taqman qPCR assays for CDK4 , CDK6 , HPV16 E6 , HPV6 E7 , HPV11 E7 , HPV16 E7 , HPV18 E7 , KDM6A , KDM6B , and p16INK4A ( supplied by Applied Biosystems as a 20× premix containing both primers and FAM-nonfluorescent quencher probe ) or SYBR Green PCR Master Mix and the listed PCR primers ( S3 Table ) to analyze expression of CDT1 , CDC7 , DBF4 , p21CIP1 , and p27KIP1 . Analysis was performed using a StepOnePlus Real-Time PCR System ( Applied Biosystems ) . Data shown are calculated using the ΔΔCT method and are normalized to expression of 18s rRNA ( Taqman ) or GAPDH ( SYBR Green ) as the housekeeping gene . ChIP was performed using the Simple ChIP Plus Enzymatic Chromatin IP Kit ( Cell Signaling ) . Immunoprecipitation of cross-linked chromatin was conducted the following antibodies: H3 ( 2560; Cell Signaling ) , H3K27me3 ( ab6002; Abcam ) , and IgG ( 2729; Cell Signaling ) . After immunoprecipitation , extracted DNA was amplified by real-time qPCR using the oligonucleotide primers described in S4 Table . Student t test was used to evaluate statistical significance . Cell lysates were prepared and processed as described [16] . Antibodies were used at the following dilutions: β-actin ( MAB1501 , 1:1 , 000; Chemicon ) , KDM6A ( ab36938 , 3μg/mL; Abcam ) , p21CIP1 ( ab109520 , 1:750 and ab109199 , 1:1 , 000; Abcam ) , FLAG ( F-3165 , 1:1 , 000; Sigma ) and HRP-conjugated secondary anti-rabbit ( 1:10 , 000; Amersham ) and anti-mouse ( 1:10 , 000; Amersham ) . Antigen/antibody complexes were visualized by enhanced chemiluminescence ( PerkinElmer Life Sciences ) and electronically acquired with a Kodak 4000R Image Station ( Kodak ) equipped with Carestream Molecular Imaging Software . Immunofluorescence analysis of monolayer cells was performed as described [16] using anti-53BP1 ( ab172580; Abcam ) and secondary donkey anti-goat Alexa Fluor 488 secondary antibody ( ab150077; Abcam ) . Nuclei were counterstained with Hoechst 33258 . Images were acquired using an Axioplan 2 microscope ( Zeiss ) with a 63× objective and Axiovision 4 . 8 ( Zeiss ) software . Unsynchronized cells were pulse labeled for 120 min with growth medium containing 10 mM of the thymidine analog iododeoxyuridine ( IdU ) . At the end of the first labeling period , the cells were washed twice with warm medium and pulse labeled once more for 30 min with growth medium containing 10 mM of the thymidine analog chlorodeoxyuridine ( CldU ) . Cells were then harvested , and genomic DNA was extracted and combed as previously described [95] . The primary antibody for fluorescence detection of IdU was mouse anti-BrdU ( Becton Dickinson ) , and the secondary antibody was goat anti-mouse mouse-Dylight 488 ( Abcam ) . The primary antibody for fluorescence detection of CldU was rat anti-CldU ( Serotec ) . The secondary antibody was goat anti-rat Alexa Cy3 ( Abcam ) .
|
High-risk human papillomaviruses ( HPVs ) are associated with approximately five percent of all human cancers , including virtually all cervical cancers as well as a large percentage of anal , vaginal , vulvar , penile , and oropharyngeal cancers . The HPV E6 and E7 proteins are the major oncogenic drivers in these tumors , and persistent expression of E6 and E7 is required for the maintenance of the transformed state . While E6 and E7 lack intrinsic enzymatic activities , and thus are difficult to directly target therapeutically , they biochemically interact with , functionally modify , or alter expression of key host cellular signaling proteins . HPV16 E7 triggers increased expression of the KDM6A histone demethylase , and KDM6A expression becomes necessary for the survival of HPV16 E7 expressing cells . Here we show that the requirement for persistent KDM6A expression is mediated by the cell cycle and DNA replication inhibitor p21CIP1 in that p21CIP1 expression is necessary for survival of E7 expressing cells . Remarkably , this is based on the ability of p21CIP1 to inhibit cellular DNA replication by binding PCNA . Our results suggest that increased KDM6A and p21CIP1 expression serves to curb HPV16 E7-induced replication stress to levels that are conducive to DNA replication but do not cause death of HPV infected cells .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"cervical",
"cancer",
"pathogens",
"cell",
"cycle",
"and",
"cell",
"division",
"carcinomas",
"cell",
"processes",
"microbiology",
"social",
"sciences",
"cancers",
"and",
"neoplasms",
"viruses",
"oncology",
"dna",
"replication",
"dna",
"viruses",
"dna",
"human",
"papillomavirus",
"papillomaviruses",
"addiction",
"medical",
"microbiology",
"hpv-16",
"gene",
"expression",
"microbial",
"pathogens",
"gynecological",
"tumors",
"biochemistry",
"psychology",
"cell",
"biology",
"nucleic",
"acids",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2017
|
KDM6A addiction of cervical carcinoma cell lines is triggered by E7 and mediated by p21CIP1 suppression of replication stress
|
Type II secretion systems ( T2SSs ) are critical for secretion of many proteins from Gram-negative bacteria . In the T2SS , the outer membrane secretin GspD forms a multimeric pore for translocation of secreted proteins . GspD and the inner membrane protein GspC interact with each other via periplasmic domains . Three different crystal structures of the homology region domain of GspC ( GspCHR ) in complex with either two or three domains of the N-terminal region of GspD from enterotoxigenic Escherichia coli show that GspCHR adopts an all-β topology . N-terminal β-strands of GspC and the N0 domain of GspD are major components of the interface between these inner and outer membrane proteins from the T2SS . The biological relevance of the observed GspC–GspD interface is shown by analysis of variant proteins in two-hybrid studies and by the effect of mutations in homologous genes on extracellular secretion and subcellular distribution of GspC in Vibrio cholerae . Substitutions of interface residues of GspD have a dramatic effect on the focal distribution of GspC in V . cholerae . These studies indicate that the GspCHR–GspDN0 interactions observed in the crystal structure are essential for T2SS function . Possible implications of our structures for the stoichiometry of the T2SS and exoprotein secretion are discussed .
Many Gram-negative bacteria use a multi-protein type II secretion system ( T2SS ) to secrete a wide variety of exoproteins from the periplasm into the extra-cellular milieu [1] , [2] , [3] , [4] . In Vibrio cholerae and enterotoxigenic Escherichia coli ( ETEC ) , cholera toxin and the closely related heat-labile enterotoxin , in addition to other virulence factors , are secreted in their folded state across the outer membrane by the T2SS [5] , [6] , [7] . The T2SSs are composed of 12 to 15 different proteins that form three distinct subassemblies: ( i ) the inner membrane platform consisting of multiple copies each of GspC , GspF , GspL and GspM with an associated cytoplasmic secretion ATPase; ( ii ) the pseudopilus , a filamentous arrangement of multiple copies of five different pseudopilins; and ( iii ) a large , pore-forming outer membrane complex , mainly consisting of the secretin GspD [8] , [9] . Secretins are multimeric outer membrane proteins composed of 50–70 kDa subunits and are among the largest outer membrane proteins known . The secretin superfamily has representatives in several other multi-protein complexes engaged in transport of large macromolecular substrates across the outer membrane [10] including the T2SS , the filamentous phage extrusion machinery [11] , the type IV pilus system ( T4PS ) [12] , [13] , [14] , and the type III secretion system ( T3SS ) [15] , [16] . Of these systems , the T2SS is most closely related to the T4PS which assembles and disassembles long filamentous fibers on bacterial surfaces and is responsible for diverse functions including attachment to host cells , biofilm formation , DNA uptake and twitching motility [17] , [18] . The T2SS secretin GspD forms a dodecameric assembly according to electron microscopy studies [19] , [20] . The C-terminal 300 to 400 residues of GspD contain the most conserved segments of the secretin superfamily , which form the actual outer membrane pore [21] , [22] , [23] . The N-terminal part of GspD consists of four domains: N0-N1-N2-N3 ( Figure 1A ) [19] , [24] . The crystal structure of the N0-N1-N2 domains of the ETEC secretin GspD has been solved previously with the assistance of a single-domain llama antibody fragment or nanobody [24] . Nanobodies are the antigen-binding fragments ( VHH ) of heavy-chain-only camelid antibodies , which have been proven as effective crystallization chaperones for challenging targets , e . g . the T2SS pseudopilins complex [25] , a trypanosomal editosome protein [26] , and activated G-protein coupled receptor [27] . In the case of the secretin GspDN0-N1-N2 structure , nanobody Nb7 provided new crystal contacts and stabilized the N0-N1 domains lobe with respect to the N2 domain . The N0 domain is structurally related to domains from several proteins in bacterial multi-protein membrane complexes [28] , [29] , [30] , [31] , and to a domain of protein gp27 from T4-related bacteriophages [32] . As expected from sequence homology , the repeat N1 and N2 domains have the same fold , whereas the N3 domain is predicted to have a similar structure [24] . The fold of the N1 domain is different from that of the N0 domain and is structurally related to the eukaryotic type I KH ( hnRNP K homology ) domain [33] . By combining crystallographic and cryo-electron microscopy studies , it has been proposed that the N0 , N1 , N2 and N3 domains form the large periplasmic vestibule of the GspD dodecamer [20] . According to a number of biochemical studies , the outer membrane protein GspD has also been reported to interact with exoproteins [20] , [34] . The inner membrane protein GspC consists of several domains: a short N-terminal cytoplasmic domain that is followed by the single transmembrane helix , a Pro-rich linker , the so-called homology region ( HR ) domain in the periplasm , a second linker and a C-terminal domain ( Figure 1A ) [35] . Most frequently , this C-terminal domain is a PDZ domain , but in some cases it is a coiled-coil domain [36] , [37] . Crystal structures of the GspC PDZ domain showed that this domain can adopt open and closed conformations [38] . It has been shown in vivo in V . cholerae that GspC and GspD interact [39] . The interaction between GspC and GspD appears critical for the function , and possibly even for the assembly , of the T2SS [39] . Besides providing a physical link between the two membranes , either or both of these proteins or their interaction could also be important for exoprotein recognition , pseudopilus formation and release of the exoprotein through the GspD pore . Biochemically , we showed that the HR domain of GspC is the key part of GspC that interacts with the periplasmic GspDN0-N1-N2 [38] . This interaction was confirmed and further investigated recently in the plant pathogen Dickeya dadantii , a species previously called Erwinia chrysanthemi [40] . The interaction between GspC and GspD of Xanthomonas campestris has also been observed in vitro [37] . We report three structures of GspCHR in complex with N-terminal domains of GspD that provide a structural basis to understand the functional interplay between the inner membrane platform and the outer membrane secretin of the T2SS . The observed interface led to the design of experiments to probe the importance of specific amino acids by biochemical and in vivo studies . Altering interface residues disabled the interaction of GspC and GspD in a bacterial two-hybrid system . It also abrogated protease secretion and had a dramatic effect on the localization of GspC in the cell envelope in V . cholerae . Together these results show the physiological importance of the molecular interactions observed between the inner and the outer platform . In addition , the resultant structure of the HR domain of GspC means that the structures of essentially all globular domains of the major T2SS proteins are presently known . The structures of ETEC GspCHR in complex with N-terminal domains of GspD reported here are the first to reveal critical interactions between the inner membrane platform and the outer membrane complex of the T2SS at the atomic level .
A complex of ETEC GspCHR and GspDN0-N1-N2 could be obtained but yielded only poorly diffracting crystals . To improve the quality of these crystals , we screened the same set of GspD specific nanobodies that had been used previously to solve the structure of GspDN0-N1-N2 [24] as crystallization chaperones for the GspCHR–GspDN0-N1-N2 complex . Using nanobody Nb3 , we obtained crystals of a ternary ETEC GspCHR–GspDN0-N1-N2–Nb3 complex , which diffracted initially only to ∼5 . 5 Å resolution . Nevertheless , a partial molecular replacement structure revealed that the HR domain of GspC interacts with the lobe formed by the N0-N1 domains of GspD . To better characterize this interaction we also crystallized smaller complexes of GspCHR–GspDN0-N1 with or without nanobodies . To assist in crystallographic phasing , we also engineered a lanthanide-binding tag ( LBT ) into the N0 domain of GspDN0-N1 [41] . The LBT to GspDN0-N1 facilitated crystal growth and the resultant crystals of the binary GspCHR–GspDN0-N1 complex diffracted to better than 2 . 7 Å resolution , with the LBT engaged in multiple crystal contacts ( Figure S1 ) . The structure of this binary GspCHR–GspDN0-N1 complex was solved by molecular replacement and refined with good crystallographic and stereochemical statistics ( Table 1 ) . In parallel , we also obtained crystals and solved the 4 Å resolution structure of a ternary GspCHR–GspDN0-N1–Nb3 complex , and also improved the diffraction of crystals of the GspCHR–GspDN0-N1-N2–Nb3 complex to ∼4 Å resolution ( Table 1 , Figure S2 ) . The three multiprotein structures obtained from different crystal forms allow a detailed description of the interactions between GspC and GpsD . In all three structures , the N0 domain of GspD interacts exclusively with the HR domain of GspC . In the 2 . 63 Å resolution binary complex , the LBT introduced into GspDN0 faces away from the interface with GspC ( Figure 1B ) . In the two low-resolution ternary complex structures , the nanobody Nb3 binds the N0 domain of GspD , opposite to the binding site of the HR domain of GspC ( Figure 1C and Figure S3 ) . In all three structures the HR domain binds in very similar orientation to GspD , relative to its N0 domain . Hence , neither the LBT nor the nanobody appears to affect the binding mode of GspC to GspD . Because the structure of the binary GspCHR–GspDN0-N1 complex has the highest resolution , this structure will be used below to analyze the specific features of the GspC–GspD interaction . The HR domain folds into a β-sandwich formed by six consecutive β-strands arranged as two three-stranded anti-parallel β-sheets ( Figure 1B ) . The residues between strands β3 and β4 adopt an approximately one-turn helical conformation . In its folded structure as seen in the complex with GspDN0-N1 ( Figure 2 ) , the distribution of charges on the surface of GspCHR is quite uneven with the main hydrophobic surface interacting with GspDN0 . Part of the remaining HR surface that is not involved in the GspD interaction ( upper panel Figure 2 ) has a preponderance of negative charges and a deep pocket defined by residues Val127 , Ile142 and Leu157 . The other side of the HR domain ( lower panel Figure 2 ) displays a mix of positive , negative and hydrophobic patches . The functions of these features during assembly and action of the T2SS , if any , remain to be determined . The closest known structural homolog of the HR domain of ETEC GspC appears to be Neisseria meningitidis lipoprotein PilP ( NmPilP ) which interacts with the secretin of the T4PS [42] . The HR domain of GspC and the core domain of NmPilP superimpose with an r . m . s . deviation of 1 . 6 Å and 25% sequence identity over 59 residues ( Figure 3 ) . The structure of NmPilP has been described as a β-sandwich composed of 7 β-strands [42] . Whereas residues 154–156 of GspC , corresponding to strand β4 of NmPilP , make some main chain hydrogen bonds to residues in strand β4 of GspC ( corresponding to β5 of NmPilP ) , the secondary structure assignment algorithm of DSSP [43] does not classify these residues as β-structure . A potential binding site has been described for the core NmPilP domain [42] . It consists of a hydrophobic crevice on the open end of the β-sandwich . The residues that create this hydrophobic groove appear to be conserved between these two proteins from the T2SS and the T4PS when they are superimposed ( Figure 3C ) . However , the area equivalent to the NmPilP pocket is covered by residues N-terminal to strand β1 in ETEC GspC and , therefore , the NmPilP pocket is absent in GspC ( Figure 3B ) . These differences do not appear to stem from crystal contacts in the GspCHR–GspDN0-N1 structure . Moreover , these residues are well conserved ( Figure S4A ) and contribute to the hydrophobic core of the HR domain . The full implications of the global structural similarity between the core PilP domain of the T4PS and the HR domain of GspC from the T2SS remain to be established , but it is in line with several known similarities between the T2SS and T4PS [17] , [18] . The interface between GspCHR and GspDN0 buries 1280 Å2 of accessible surface area with a calculated ΔG of interaction of −5 . 4 kcal×mol−1 as assessed by the PISA server ( Figure 4 ) [44] . The overall shape of the interface is relatively flat with a small concave area on the GspD surface . A total of 18 residues from GspCHR and 19 residues from GspDN0 engage in a combination of hydrophobic interactions and hydrogen bonds . The first three β strands of GspCHR and the first β strand plus the subsequent helix α1 of GspDN0 are the major contributors to the interface . The majority of the hydrogen bonds are formed by an antiparallel arrangement of strand β1 of GspCHR and strand β1 of GspDN0 ( Figure 4C ) . This β-strand augmentation is frequently observed in protein–protein interfaces [45] . Several nonpolar residues are engaged in intermolecular hydrophobic interactions , e . g . Ala133/Val141 from GspC , and Phe5/Phe9 from GspD . The hydrophobic nature of these interacting residues is well conserved , with GspC residue 133 being Ala , Leu , Val or Met; GspC residue 141 either a Val or Ile; GspD residue 5 a Phe or Tyr; and GspD residue 9 a Phe according to a family sequence alignment ( Figure S4 ) . Nonetheless , the GspC–GspD interface provides a species-specific connection point between outer and inner membrane assemblies of the T2SS as has been observed in genetic complementation studies [46] , [47] . Based on the GspCHR–GspDN0-N1 structure , we selected several well-conserved interface residues for subsequent substitutional analysis . Ala133 and Val141 from ETEC GspC ( equivalent to Val118 and Val129 from V . cholerae GspC , respectively; Figure 4E ) and Thr20 from ETEC GspD ( equivalent to Ile18 of V . cholerae GspD ) are completely buried upon complex formation and are located in the center of the interacting surfaces ( Figure 4D ) . Asn24 from ETEC GspD ( equivalent to Asn22 of V . cholerae GspD ) makes a hydrogen bond with the main chain oxygen of ETEC GspC Arg137 . We evaluated the role of these residues on complex formation of truncated forms of GspC and GspD in a bacterial two-hybrid system and in a functional V . cholerae secretion assay in vivo . We also assessed the effect of interface substitutions on the distribution of GspC in the cell envelope of V . cholerae . The effect of several interface substitutions on the ability of GspD to associate with GspC was assayed in a bacterial two-hybrid system based on reconstitution of activity of the catalytic domain of Bordetella pertussis adenylate cyclase when T18 and T25 fragments are fused to interacting proteins ( see Methods ) [48] . VcGspD–T18 with a conservative Asn22Gln substitution retained the ability to interact with T25–VcGspC and formed dark red colonies on indicator agar . In contrast , VcGspD–T18 with either an Asn22Arg substitution or an Ile18Asp substitution lost the ability to interact with T25–VcGspC and formed colorless colonies ( Table 2 ) . Two variants of T25–VcGspC , with either Val118Arg or a Val129Arg substitution , also lost the ability to interact with VcGspD–T18 and formed colorless colonies in the bacterial two-hybrid system . The functional importance of residues involved in the GspC–GspD interface was also assessed in vivo by monitoring the effect of the Ile18Arg and Asn22Tyr mutations in VcGspD on the extracellular secretion of protease by V . cholerae . No protease secretion was observed when plasmid-encoded VcGspDIle18Arg/Asn22Tyr was produced in a V . cholerae mutant strain lacking the gene encoding VcGspD ( Figure 5A ) , indicating that the simultaneous exchange of these two amino acids prevents protein secretion by the T2SS . The singly substituted variants , however , remained functional ( Figure 5A ) . Immunoblot analysis of cell extracts from the ΔgspD mutant strain producing plasmid-encoded wild type and mutant VcGspD showed that the double VcGspD mutant protein was made at levels similar to that of wild-type VcGspD ( Figure 5B ) . Using V . cholerae strains producing chromosomally encoded VcGspC fused to the green fluorescent protein ( GFP ) , we visually examined the effects of substitutions in the GspC–GspD interface on subcellular localization of GspC . GFP-VcGspC forms fluorescent foci in the V . cholerae cell envelope , which disperse upon deletion of the gene encoding VcGspD and reappear when the deletion strain is complemented with plasmid-encoded VcGspD ( Figure 6 , first and second panels ) [39] . The substitution of wild-type VcGspD with VcGspDIle18Arg/Asn22Tyr resulted in loss of fluorescent foci and dispersal of the fluorescence in a manner indistinguishable from cells that do not have the gene encoding VcGspD at all ( Figure 6 , fourth panel ) . This result suggests that residues Ile18 and Asn22 of VcGspD are critical for the incorporation of GFP-VcGspC fusion protein into fluorescent foci , and supports the suggestion that the interaction between GspC and GspD observed in the crystal structure of GspCHR in complex with GspDN0-N1 ( Figure 4 ) is physiologically relevant . Based on these results , it appears that VcGspD has to interact directly with VcGspC in order to support its focal distribution in V . cholerae .
The structures of the first two domains of related secretins have been determined in two prior studies: ETEC GspD from the T2SS and EPEC EscC from the T3SS [24] , [49] . The relative orientations of the N0 and N1 domains in these two studies appeared to be remarkably different: when the N1 domains of the T2SS and T3SS secretins are superimposed , the N0 domains are rotated by not less than 143 degrees [10] . This raises an important question as to the actual orientation of these two domains in the T2SS and T3SS secretins . Regarding the T2SS , the relative orientations of the N0 and N1 GspD domains can now be compared in two high resolution structures , i . e . in the current structure of the binary complex of ETEC GspCHR and GspDN0-N1 , and in the previously determined binary complex of ETEC GspDN0-N1-N2 in complex with Nb7 [24] . The linker between the N0 and N1 subdomains is disordered in both these high resolution structures . The interface and relative orientation of the N0 and N1 subdomains , however , is essentially the same in the two structures despite the binding of either Nb7 or the presence of the LBT insertion into the N0 domain: the superposition of the two N0 domains results in an r . m . s . deviation of 0 . 49 Å for 72 Cα pairs ( Figure S5 ) . Taking also into account the two new low resolution structures of the ternary complexes of GspCHR–GspDN0-N1–Nb3 and GspCHR–GspDN0-N1-N2–Nb3 ( Figure S3 ) , then the N0-N1 lobe in the T2SS secretin GspD is observed as the same compact unit in four different crystal structures , independent of the presence or absence of a GspCHR domain , Nb molecules or crystal contacts . The available data suggest that the N0-N1 orientation in GspD is a characteristic feature in the T2SS . However , we cannot exclude the possibility that the relative orientation of the N0 and N1 domains may change as the secretin oligomerizes . Only high resolution structures of the dodecameric secretin will resolve this question . Since the N0-N1 lobe of the T2SS secretin fits well into the cryo-electron microscopy reconstruction of VcGspD [20] , and the N0 and N1 domains of the T3SS secretin fit well into a cryo-electron microscopy density of the Salmonella typhimurium needle complex [16] , it might be that the N0 and N1 domains of these related secretins adopt different mutual orientations in the assembled T2SS and T3SS in vivo as observed in crystals . Obviously further studies are required to confirm this hypothesis where it also should be kept in mind that secretins are dynamic proteins and multiple orientations of N-terminal secretin domains might transiently occur during the secretion process [10] . The crystal structure indicates that a number of residues are critical for the interactions of ETEC GspC and GspD ( Figure 4 ) . Moreover , these residues are conserved in the family sequence alignment ( Figure S4 ) . As many mutants and other useful reagents have already been generated and developed for studies of the T2SS in V . cholerae , subsequent probing of the importance of these residues for the interaction was carried out in three different ways using V . cholerae GspC and GspD homologues . The two-hybrid studies showed that substitutions Val118Arg and Val129Arg in VcGspC , and Asn22Arg in VcGspD , abrogated the interaction between GspCHR and GspDN0-N1-N2 from V . cholerae ( Table 2 ) . The secretion of protease by V . cholerae was also greatly diminished by substitutions Ile18Arg/Asn22Tyr in full length VcGspD ( Figure 5 ) . Finally , the same Ile18Arg/Asn22Tyr variant of VcGspD altered the distribution of full-length VcGspC in the inner membrane of V . cholerae , possibly by interfering with normal assembly of the inner membrane platform of the T2SS ( Figure 6 ) . Taking all data together , we conclude that the substitutions altering the interface of GspC with GspD in V . cholerae affect the interactions of GspC with GspD as demonstrated both in a bacterial two-hybrid system and by analysis of protease secretion by the T2SS in V . cholerae . Interactions between GspC and GspD from D . dadantii have been recently investigated [40] . This study confirmed the interactions between GspCHR and the N-terminal domains of GspD reported earlier for V . vulnificus homologs [38] . A GST-fusion of residues 139–158 of DdGspC ( corresponding to residues 168–187 in ETEC GspC ) co-purified with both DdGspDN0 and DdGspDN1-N2-N3 [40] . The 139–158 residues of DdGspC were therefore designated as secretin interacting peptide ( SIP ) . In a homology model of DdGspCHR–GspDN0-N1 complex , based on our crystal structure , this fragment is located far from the interface ( Figure S6 ) . It appears that this segment forms an anti-parallel pair of β-strands , β5 and β6 , in the ETEC GspCHR crystal structure , with β6 at the surface and β5 located between strands β6 and β4 ( Figure S6 ) . Furthermore , the substitutions introduced into the DdGspC 139–158 fragment had no effect on the interaction with DdGspDN0 , whereas one substitution , Val143Ser , prevented DdGspC interaction with DdGspDN1-N2-N3 [40] . The same substitution , when introduced into full length DdGspC , also interfered with secretion in D . dadantii . We also mapped these substitutions onto the homology model of the DdGspC–GspD complex and it is clear that none of them are buried in the GspC–GspD interface ( Figure S6 ) . The only substitution that had an effect on secretion , Val143Ser , replaces a buried hydrophobic residue in the core of DdGspCHR with a polar residue that would likely be detrimental to the HR domain stability . This is in agreement with the finding that this substitution in GST-DdGspC128-272 resulted in a protein that is degraded in the cells [40] . A more conservative Val143Ala substitution in full length DdGspC appeared to largely support secretion of pectinases , in agreement with the less drastic change of the nature of the side chain , which could result in a larger proportion of properly folded protein than in the case of the Val143Ser variant . Therefore , the ETEC GspCHR–GspDN0-N1 structure explains several experimental results of the studies on DdGspC–GspD interactions [40] . The observations that a GST-fusion of the DdGspC 139–158 fragment is capable of interacting with fragments of the secretin in the absence of both the rest of the HR domain and the rest of the secretin , and of interfering with pectinase secretion when over-expressed in wild type D . dadantii , are difficult to interpret precisely . Additional studies are required to show that such interactions are not the result of non-specific interactions , possibly due to exposed hydrophobic residues of the peptide which are normally buried in the complete HR domain . The implications of the GspCHR–GspDN0 interactions unraveled by our studies for the architecture of the T2SS are intriguing . The three new structures in the current paper show that one GspCHR domain interacts with one GspDN0 domain , which suggests a 1∶1 ratio of GspC and GspD in the assembled T2SS . Since the stoichiometry of full length GspC and GspD has not been established yet in the context of a functional T2SS , it is of interest to see if the current complex of GspCHR–GspDN0 is compatible with the dodecameric ring of GspDN0-N1 derived recently by a combination of crystallographic and electron microscopy studies [20] , [24] . Superimposing the GspCHR–GspDN0 complex twelve times onto the N0-domains of the GspDN0-N1 ring results in a double ring structure where the GspCHR subunits added do not interfere with the formation of the GspDN0-N1 ring . Although this procedure does result in some clashes between the subunits of the GspCHR ring , specifically between residues of the β2-β3 loop of one subunit and residues just prior to β6 in a neighboring subunit , small conformational changes in these loops , or minor adjustments in the mutual orientation of domains in the GspDN0 ring , or both , might alleviate these close contacts . If this would be the case , the GspD dodecamer would interact with twelve GspC subunits in the assembled T2SS ( Figure 7A ) . Alternatively , only alternating GspD subunits of the dodecameric secretin might interact with GspCHR , obviously removing close contacts between the then well separated GspCHR subunits . In this case , the GspD dodecamer would interact with six GspC subunits ( Figure 7B ) . These two alternatives for the interface of the outer membrane complex and the inner membrane platform can be combined with previous studies on the T2SS even though the ratio between GspC and the other components of the inner-membrane platform complex is currently unknown . Yet , the following observations are of interest for the T2SS stoichiometry puzzle: ( i ) the secretion ATPase GspE of the T2SS has been suggested to be a hexamer [50] , [51]; ( ii ) the cytoplasmic domain of the inner membrane T2SS protein GspL forms a 1∶1 complex with GspE [52]; ( iii ) homologs of GspM and of the cytoplasmic domain of GspL from the T4PS have been reported to form heterodimers [53] , [54]; ( iv ) there are a few cases of gene fusion of the T4PS proteins PilP and PilO ( e . g . Pseudomonas putida PilO-PilP , Uniprot entry Q88CU9 ) in the T4PS . PilP is a GspCHR homolog ( Figure 3 ) and PilO is proposed to be a homolog of the inner membrane protein GspM from the T2SS [54] , [55] . The presence of PilO–PilP fusions may imply a 1∶1 stoichiometry of these proteins in the T4PS and , in view of the homology between the T4PS and the T2SS , a GspM:GspC ratio of 1∶1 in the T2SS as well . These four observations suggest that GspE , GspL , GspM and GspC might be present in an equimolar ratio in the inner membrane platform . In view of the likely hexameric nature of GspE , this implies the presence of six subunits of each of these proteins in the assembled T2SS . If the GspD dodecamer would interact with six GspC subunits ( Figure 7B ) , then this arrangement would agree well with six subunits each of GspC , GspL , GspM and GspE in the inner membrane platform . If a GspD dodecamer , however , would interact with twelve GspC subunits in the assembled T2SS ( Figure 7A ) , then , a stoichiometry mismatch is likely to occur somewhere along the GspC–GspL–GspM–GspE chain of interactions in the inner membrane platform . This could be possible in spite of the evidence in points ( i ) to ( iv ) above for an equimolar ratio of these four proteins in the T2SS since e . g . points ( iii ) and ( iv ) are rather indirect and derived from observations on T4PS proteins . Clearly , the stoichiometry of the T2SS remains a fascinating topic for further studies , where the number of GspF subunits , the only T2SS protein which spans the inner membrane multiple times , also remains to be determined . Another major outstanding question is the recognition of exoproteins by the T2SS . Interestingly , the inner diameter of the dodecameric GspCHR–GspDN0-N1 double ring is ∼68 Å , which implies that a large exoprotein like the cholera toxin AB5 heterohexamer [56] just fits into this ring ( Figure 7C ) . This is in agreement with recent electron microscopy studies which indicate that the B-pentamer of cholera toxin can bind to the entrance of the GspD periplasmic vestibule [57] . The periplasmic domains of GspD and of GspC have been implicated in this crucial exoprotein recognition function [34] , [46] , [57] , [58] , [59] , but the specific details of exoprotein–T2SS interactions remain to be uncovered . The accumulation of recent structural and biochemical data provides a platform for asking increasingly precise questions regarding the many remaining mysteries still pertaining to the architecture and mechanism of the sophisticated T2SS .
ETEC GspDN0-N1-N2 ( residues 1–237; numbering corresponds to mature protein sequence ) was expressed and purified as described [24] . The DNA sequence corresponding to residues 1–165 of ETEC GspD was PCR amplified and cloned into the pCDF-NT vector to obtain a GspDN0-N1 expression construct . pCDF-NT is a modified pCDF-Duet1 vector ( Novagen ) encoding an N-terminal His6-tag sequence and a TEV protease cleavage site . The DNA sequence corresponding to residues 122–186 of ETEC GspC was PCR amplified and cloned into a pCDF-NT vector to obtain a GspCHR expression construct . A lanthanide binding tag ( LBT ) was introduced into GspDN0-N1 construct in order to assist with crystallographic phase determination and promote crystal formation . In order to decrease the flexibility of the LBT , we introduced it into the loop between two adjacent β-strands rather than at the termini . The design was based on the crystal structure of ubiquitin with the double LBT ( PDB 2OJR ) [41] where two β-strands flank one of the LBT . The LBT sequence YIDTNNDGYIEGDEL was inserted between residues Met70 and Val74 of GspDN0 ( Figure S4B ) using the polymerase incomplete primer extension method [60] . While this manuscript was in preparation , a similar approach for the LBT insertion was successfully applied to a model protein , interleukin-1β [61] . GspDN0-N1 was expressed at 25°C in BL21 ( DE3 ) cells ( Novagen ) in LB medium containing 50 µg×ml–1 streptomycin . Protein production was induced with 0 . 5 mM IPTG . Cells were harvested 3 h after induction . GspDN0-N1 variants with or without LBT were purified by Ni-NTA agarose ( Qiagen ) chromatography followed by His6-tag cleavage with TEV protease; a second Ni-NTA chromatography to remove His6-tag , uncleaved protein and His-tagged TEV protease; and a final size-exclusion chromatography using Superdex 75 column ( GE Healthcare ) . GspCHR was expressed and purified under same conditions as GspDN0-N1 . The proteins were concentrated , flash-frozen [62] and stored at −80°C . Se-Met-labeled proteins were expressed using metabolic inhibition of methionine biosynthesis [63] and purified using the protocols for native proteins . The nine nanobodies generated against ETEC GspDN0-N1-N2 were expressed and purified as described previously [24] . ETEC GspCHR , GspDN0-N1-N2 and individual nanobodies were mixed at 1∶1∶1 molar ratio , concentrated to 4–8 mg×ml−1 total protein concentration and subjected to crystallization conditions screening by the vapor diffusion method at 4 or 21°C . The crystallization conditions were identified using SaltRx ( Hampton Research ) and JCSG+ ( Qiagen ) screens . The complex of GspCHR–GspDN0-N1-N2–Nb3 was crystallized in 1 . 2 M lithium sulfate , 0 . 1 M Tris-HCl pH 7 at 4°C . The crystals were gradually transferred to precipitant solution supplemented with 30% glycerol and flash-frozen in liquid nitrogen . Initial crystals diffracted to 5 . 5 Å resolution and optimized crystals with Se-Met substituted GspDN0-N1-N2 showed improved diffraction to 4 . 6 Å . Data were processed and scaled using XDS [64] . The structure was solved by molecular replacement using Phaser [65]; the search models included the GspDN0-N1 structure ( PDB 3EZJ ) [24] , a camelid antibody fragment ( PDB 1QD0 ) [66] , and a homology model of GspCHR obtained using the I-TASSER server [67] and the N . meningitidis PilP structure as template ( PDB 2IVW ) [42] . The N2 domain of GspD could not be located in the electron density maps due to statistical disorder ( Figure S2 ) . The complex of GspCHR–GspDN0-N1 with an engineered LBT in the GspDN0 domain was crystallized in 0 . 9 M magnesium sulfate , 0 . 1 M bis-tris propane pH 7 . 0 at 21°C . The crystals were transferred to precipitant solution supplemented with 20% ethylene glycol and flash-frozen in liquid nitrogen . The structure of the GspCHR–GspDN0-N1 complex was solved by molecular replacement using Phaser and rebuilt using Buccaneer [68] and Coot [69] . The metal binding site of the LBT appears to be occupied by a Ca2+ ion based on the electron density and the B factor values after refinement ( Figure S1 ) . Most likely , Ca2+ ions were acquired during E . coli expression , which prevented Tb3+ ions binding during treatment of purified protein according to a published protocol [41] . The capture of ions during heterologous expression by Ca2+ binding proteins has been observed previously for the major pseudopilin GspG [70] . The structure was refined with REFMAC [71] using translation , libration and screw-rotation displacement ( TLS ) groups defined by the TLSMD server [72] . The quality of the structure was assessed using the Molprobity server [73] . The ternary GspCHR–GspDN0-N1–Nb3 complex was crystallized in 0 . 7 M sodium citrate , 0 . 1 M bis-tris propane , pH 7 . 0 at 21°C . The crystals were cryoprotected using 20% ethylene glycol . The structure of the GspCHR–GspDN0-N1–Nb3 complex was solved by molecular replacement using Phaser with refined GspCHR and GspDN0-N1 fragments from our GspCHR–GspDN0-N1 structure as search models ( Figure S3 ) . Protein–protein interfaces were evaluated using the PISA server [44]; structural homologs were searched for using the DALI server [74]; the electrostatic surface potential was calculated using APBS [75]; figures were prepared using PyMol [76] . Interaction between protein domains was detected by the ability of fusion proteins containing the enzymatically inactive T18 and T25 fragments of adenylate cyclase toxin from Bordetella pertussis to confer adenylate cyclase activity ( and the ability to ferment maltose and form red colonies on maltose-MacConkey plates ) to a cyaA mutant E . coli strain as described previously [48] . E . coli DC8F′ is a cyaA::KmR derivative of the strain MM294 ( endA1 hsdR17 glnV44 thi-1 ) with the TcR F′ lacIq Tn10 from XL1blue ( Stratagene ) . Plasmids pCT25VcGspC ( encoding a T25–VcGspC fusion protein ) and pAVcGspDT18 ( encoding a VcGspD–T18 fusion protein ) were separately transformed into E . coli DC8F′ , and transformants were selected on LB-Cm or LB-Ap plates , respectively . Each of the resulting transformants formed white colonies when streaked onto maltose MacConkey plates and incubated at 30°C . In contrast , when both plasmids were transformed together into E . coli DC8F′ , the resulting transformants formed red colonies when streaked onto maltose MacConkey plates , demonstrating a productive protein-protein interaction between the VcGspC and VcGspD domains of the T25–VcGspC and VcGspD–T18 fusion proteins , bringing together the T18 and T25 fragments to form active cyclase . A positive control also demonstrated a productive protein-protein interaction between CTA1R7KT18 and CT25ARF6 fusion proteins in E . coli DC8F′ and formation of red colonies on maltose MacConkey agar , as reported previously [48] . Negative controls failed to demonstrate any productive protein-protein interaction between the CTA1R7KT18 and T25VcGspC fusion proteins or between theVcGspDT18 and CT25ARF6 fusion proteins in E . coli DC8F′ . A DNA sequence encoding residues 53–305 of VcGspC ( AAA58784 . 1 ) was amplified by PCR using the primers EpsCXF and EpsCHIIIR adding XbaI ( Leu-Glu frame ) and a stop codon-HindIII sites at the 5′ and 3′ ends respectively . This product was cloned in frame after the T25 domain in place of ARF6 in pXCT25arf6 ( pCT25ARF6 from [48] but with a vector XbaI site deleted ) to generate pCT25VcGspC . Similarly the coding sequence for residues 25–294 of VcGspD ( AAA58785 . 1 ) was amplified with primers EpsDNdeIF and EpsDClaR which add NdeI ( and Met codon ) and ClaI ( Ser-Met frame ) sites at the 5′ and 3′ ends respectively; this PCR product was cloned in place of the CTA1 gene in pCTA1R7KT18 [48] to generate pAVcGspDT18 . The primers sequence information is available upon request . Specific mutations in the eps gene domains ( encoding GspC or GspD ) in pAVcGspDT18 and pCT25VcGspC were generated by SOE-PCR [77] or by subcloning of a PCR fragment performed with a restriction site containing mutagenic primer and a vector primer , followed by recloning into the parental vector . All clones were verified in-frame and correct by DNA sequencing to ensure no additional PCR-generated mutations . The ΔgpsD strain of V . cholerae , a gfp-gspC ΔgspD strain , and the complementing pMMB-gspD plasmid were constructed previously [39] . Mutations were introduced in the gspD gene of V . cholerae with the QuikChange II site-directed mutagenesis kit ( Stratagene ) using pBAD-gspD as a template . Primers used for the site change in gspDI18R and gspDN22Y were 5′-GAATTTATCAATCGTGTGGGACGCAATC-3′ , 5′-GATTGCGTCCCACACGATTGATAAATTC-3′ and their reverse complements , respectively . gspDI18R/N22Y was then constructed using pBAD-gspDI18R as a template and the above primers specific for the gspDN22Y site change . All mutations were verified by sequencing . Following sequencing , the gspD variants of V . cholerae obtained were cloned into the low-copy-vector pMMB67 using restriction enzymes BamHI and SphI . V . cholerae cultures were grown overnight at 37°C in Luria broth supplemented with 100 µg×ml−1 thymine , 200 µg×ml−1 carbenicillin , and 20 µM IPTG and centrifuged to separate the supernatant and cellular material . The supernatants were centrifuged once more , and the protease activity was measured as described previously [78] . Cultures of V . cholerae were grown overnight at 37°C in M9 medium containing 0 . 4% casamino acids , 0 . 4% glucose , and 100 µg×ml−1 thymine; diluted 50–fold into fresh medium; and grown to mid-log phase before observation . Plasmids were maintained with 50 and 200 µg×ml−1 carbenicillin in log-phase and overnight cultures , respectively . Plasmid expression was induced with IPTG as described above . For fluorescence microscopy of live cells , cultures were mounted on 1 . 5% low-melting temperature agarose pads prepared with M9 glucose medium . All microscopy was performed with a Nikon Eclipse 90i fluorescence microscope equipped with a Nikon Plan Apo VC100 ( 1 . 4 numerical aperture ) oil immersion objective and a Cool SNAP HQ2 digital camera . Captured images were analyzed with NIS-Elements imaging software ( Nikon ) . Atomic coordinates and structure factors have been deposited in the Protein Data Bank ( http://www . pdb . org ) with accession code 3OSS .
|
Many bacterial pathogens affecting humans , animals and plants export diverse proteins across the cell membranes into the medium surrounding the bacteria . Some of these secreted proteins are involved in pathogenesis . One example is cholera toxin secreted by the bacterium Vibrio cholerae , a causative agent of cholera . The sophisticated type II secretion system is responsible for moving this toxin , and several other proteins , across the outer membrane . Here , we studied the interaction between the outer membrane pore of the type II secretion system , the secretin GspD , and the inner membrane protein GspC . We have solved three crystal structures of complexes between the interacting domains and identified critical contacts in the GspC–GspD interface . We also showed the importance of these contacts for assembly of the secretion system and for secretion of proteins by V . cholerae . Our studies provide a major piece in the puzzle of how the type II secretion system is assembled and how it functions . One day this knowledge might allow us to design compounds which interfere with this secretion process . Such compounds would be useful in the battle against bacteria affecting human health .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"biochemistry",
"gram",
"negative",
"transmembrane",
"proteins",
"protein",
"interactions",
"proteins",
"protein",
"structure",
"macromolecular",
"assemblies",
"microbial",
"pathogens",
"biology",
"transmembrane",
"transport",
"proteins",
"microbiology",
"bacterial",
"pathogens",
"pathogenesis"
] |
2011
|
Structural and Functional Studies on the Interaction of GspC and GspD in the Type II Secretion System
|
Chagas disease , caused by the protozoan parasite Trypanosoma cruzi ( T . cruzi ) , is a life threatening global health problem with only two drugs available for treatment ( benznidazole and nifurtimox ) , both having variable efficacy in the chronic stage of the disease and high rates of adverse drug reactions . Inhibitors of sterol 14α-demethylase ( CYP51 ) have proven effective against T . cruzi in vitro and in vivo in animal models of Chagas disease . Consequently two azole inhibitors of CYP51 ( posaconazole and ravuconazole ) have recently entered clinical development by the Drugs for Neglected Diseases initiative . Further new drug treatments for this disease are however still urgently required , particularly having a different mode of action to CYP51 in order to balance the overall risk in the drug discovery portfolio . This need has now been further strengthened by the very recent reports of treatment failure in the clinic for both posaconazole and ravuconazole . To this end and to prevent enrichment of drug candidates against a single target , there is a clear need for a robust high throughput assay for CYP51 inhibition in order to evaluate compounds active against T . cruzi arising from phenotypic screens . A high throughput fluorescence based functional assay using recombinantly expressed T . cruzi CYP51 ( Tulahuen strain ) is presented here that meets this requirement . This assay has proved valuable in prioritising medicinal chemistry resource on only those T . cruzi active series arising from a phenotypic screening campaign where it is clear that the predominant mode of action is likely not via inhibition of CYP51 .
Chagas disease is a tropical parasitic disease caused by the flagellate eukaryotic ( protozoan ) parasite Trypanosoma cruzi ( T . cruzi ) , endemic in Latin America and now emerging in North America and Europe through human migration . It is becoming a severe global health problem with approximately 8–10 million people infected , an estimated 12 , 000 deaths per year , and placing 100 million people at risk . Transmission to humans and other mammals is predominantly by an insect vector , the blood-sucking "kissing bugs" of the subfamily Triatominae ( family Reduviidae ) [1] . Transmission has also been reported to occur through contaminated food , blood transfusions and from mother to child . Clinical Chagas disease can be classified into two distinct phases , acute and chronic . In the acute phase , lasting a few weeks , parasites begin to multiply in the organs and tissues . Symptoms are usually mild and non-specific with patients rarely being diagnosed . However , life-threatening myocarditis or meningoencephalitis can occur during the acute phase with a death rate for people in this phase of about ten percent . Ten to fifty percent of infected survivors develop chronic Chagas disease . People in the chronic phase can be asymptomatic for many years , with parasites generally undetectable in the blood . However , the disease causes organ and tissue damage , particularly potentially lethal cardiopathy and megacolon or megaoesophagus , caused by the sequential induction of inflammatory response to the parasite . Nitroheterocyclic compounds , benznidazole and nifurtimox , developed in the 1960’s [2] , are currently the only two drugs used for the treatment of Chagas disease . Both have low efficacy in the chronic stage and , with prolonged dosing regimens , both drugs have significant side effects including skin irritation , neurotoxicity , and digestive system disorders [3] . Newer , safer and more efficacious treatments are therefore in desperate need . Inhibition of sterol 14α-demethylase ( CYP51 ) has been considered a viable target against T . cruzi for over 30 years [2 , 4 , 5 , 6 , 7 , 8] . Found in a broad variety of organisms including animals , plants , fungi and protozoa , this enzyme plays an essential role in the sterol biosynthetic pathway , catalysing the oxidative removal of the 14α-methyl group from sterol precursors such as lanosterol or eburicol [9] . The products of the pathway , cholesterol in humans or ergosterol in fungi , are required for the integrity of the eukaryotic cell membrane . These sterols are required for membrane function in T . cruzi . Inhibition of CYP51 activity is lethal as the T . cruzi parasites are unable to scavenge and utilise host cholesterol [10] . The CYP51 gene is known to be expressed in all stages of the parasite life cycle and indeed it has also been shown to be up-regulated in multiplying forms [9] . As with other members of the Cytochrome family , CYP51 is a haem containing protein located on the membrane of the endoplasmic reticulum that relies upon electron transfer by NADPH reductase for activation [11] . Azole inhibitors , which interfere with sterol biosynthesis , essential in eukaryotic cells , have already been used with success in humans in the treatment of fungal infections . Several of these drugs have been considered as possible treatments for Chagas disease [12] . Ketoconazole , fluconazole , itraconazole , ravuconazole and posaconazole are known to inhibit CYP51 in vitro , competitively binding to the haem within CYP51 and occupying the active site preventing any substrate from binding . Although ketoconazole and itraconazole have not demonstrated significant curative activity in humans with chronic Chagas disease [6] , other azoles , with greater potency and improved pharmacokinetic properties , which have been shown to have potent activity against T . cruzi , including posaconazole [13 , 14] and ravuconazole ( Fig 1 ) , are in clinical development with the Drugs for Neglected Diseases initiative ( DNDi ) . To prevent enrichment of candidates against a single target , and thus reduce risk in the overall drug discovery portfolio for Chagas disease , it has therefore become necessary to evaluate and prioritise medicinal chemistry resource on new chemical series active against T . cruzi but with such activity not likely driven via T . cruzi CYP51 inhibition . Recent findings from clinical trials with posaconazole [15] and ravuconazole [16] has indicated re-emergence of parasitaemia in two thirds of patients once dosing has been completed , thus reinforcing the need to strengthen the overall drug discovery portfolio for Chagas disease with new chemical lead series not working via this mechanism of action . Evaluating compounds as potential inhibitors of T . cruzi CYP51 has previously been demonstrated measuring the apparent dissociation constants ( Kd ) by spectral titration [4 , 17 , 18] utilising the shift of the haem iron soret band in response to binding [18] . One of the drawbacks to this methodology is that micromolar protein concentrations are required for screening causing potential interference with the optical properties and/or solubility of test compounds [4] . There are many potential reasons why affinity estimates measured by binding may not correlate with functional inhibition . These include allosteric sites , non or uncompetitive modes of inhibition or slow kinetics [19] . Inhibition of endogenous substrate lanosterol , eburicol and obtusifoliol has also been used as an in vitro tool using recombinant expressed human CYP51 enzyme [20] . In particular , measuring effect on CYP51 driven metabolism of lanosterol to follicular fluid meiosis activating sterol ( FF-MAS ) in the presence of test substances [21] is well established ( Fig 2 ) . However , FF-MAS detection requires mass spectrometry limiting the number of compounds that can be tested and consequently limiting the value of such an assay for triaging large numbers of phenotypic screening T . cruzi hits toward identifying modes of action away from CYP51 . Metabolism of fluorogenic probe substrate to a product , detectable by fluorescence is well established with recombinantly expressed cytochrome P450 enzymes ( CYP’s ) for the purpose of assessing possible drug-drug interactions [22] . Measuring CYP inhibition by this method provides a high throughput screening approach , avoiding time consuming analysis by mass spectrometry and minimising use of expensive substrates . The O-dealkylation of Vivid substrate benzyloxymethylocyanocoumarin ( BOMCC ) to fluorescent product cyanohydroxycoumarin ( CHC ) is commonly used to evaluate CYP3A4 activity in recombinantly expressed membrane preparations ( Fig 2 ) . Valuably , O-dealkylation activity in the presence of recombinantly expressed T . cruzi CYP51 was observed . This has enabled the creation of a fast , high-throughput , 96 and 384 well microtitre method to assess the inhibitory potential of compounds against T . cruzi CYP51 , which is described in this paper .
Benzyloxymethyloxycyanocoumarin ( BOMCC ) was obtained from ThermoFisher Scientific . Fluconazole , ketoconazole , itraconazole , NADP , NADPH , glucose-6-phosphate , glucose-6-phosphate dehydrogenase , Cytochrome C from horse heart and sodium bicarbonate were obtained from Sigma Aldrich . 50 mM potassium phosphate ( pH 7 . 4 ) buffer was prepared from dibasic and monobasic forms of potassium phosphate obtained from Sigma Aldrich . E . coli membrane fractions containing the T . cruzi CYP51 ( Tulahuen strain ) ( Bactosomes ) were provided by Cypex Ltd . The cDNAs coding for T . cruzi CYP51 and NADPH P450 reductase were synthesized and supplied in pUC vectors by Genescript . The CYP51 cDNA was cloned into the expression vector pCW with an ompA N terminal leader and the reductase was cloned into a pACYC184 derived expression vector with a pelB N terminal leader . E . coli JM109 was used for the co-expression of the recombinant proteins . For each concentration of test compound the rate of fluorescence units per minute was calculated as a percentage of the average rate of the solvent only control wells . The percentage of solvent control values was then plotted against the concentration range . Using the following 4 parameter fit equation , an IC50 value can be determined . Accession number for T . cruzi cDNA is AY283022; T . cruzi reductase is DQ857724 and mosquito reductase is AY183375 .
Despite the presence of the N terminal pelB leader sequence , the recombinant T . cruzi NADPH P450 reductase proved to be a cytosolic protein ( as might be expected from previous data [10] ) and was not present in the E . coli membrane fraction with the T . cruzi CYP51 . Reconstitution of the cytosolic fraction containing the T . cruzi reductase with the membrane fraction containing the T . cruzi CYP51 did not result in active CYP51 so , in order to generate an active T . cruzi CYP51 system , the T . cruzi CYP51 was co-expressed with human ( expression construct supplied by Cypex Ltd ) or mosquito ( expression construct supplied by the Liverpool School of Tropical Medicine ) NADPH P450 reductase . Screening of CYP51 activity showed that the T . cruzi CYP51 co-expressed with the mosquito reductase gave the highest rate of BOMCC turnover . The mosquito reductase expression cassette was cloned from the pACYC vector into the pCW based CYP51 expression vector to allow both proteins to be expressed from the same plasmid resulting in a relatively higher level of mosquito reductase in the membrane fraction with a concomitant increase in BOMCC turnover . Supplementation of the T . cruzi CYP51 / mosquito NADPH P450 reductase bactosomes with partially purified mosquito cytochrome b5 at 10 fold excess over the CYP51 resulted in a further increase in BOMCC turnover . This bactosome preparation was then scaled up for all further work . To determine kinetic parameters , 100 μM BOMCC O-dealkylase activity was assessed in incubations containing 25 , 50 , 75 , 100 , 150 , 200 , 250 , 300 , 400 , 500 and 600 pmoles/mL T . cruzi CYP51 enzyme . Protein concentrations were normalised using control protein containing no CYP activity . Incubations were pre-warmed at 37°C before addition of NADPH regenerating system . Production of CHC was measured at 1 minute intervals at 37°C ( Exc 410 nm , Em 460 nm ) for 10 minutes ( Fig 3a ) . Km and Vmax parameters for BOMCC were determined by incubating 37 pmoles/mL T . cruzi CYP51 enzyme ( 0 . 24 mg/mL bactosomes ) with 25 , 50 , 75 , 100 , 150 , 200 and 250 μM BOMCC . Incubations were pre-warmed at 37°C before addition of NADPH regenerating system . Production of CHC was measured at 1 minute intervals at 37°C ( Exc 410 nm , Em 460 nm ) for 10 minutes ( Fig 3b ) . Rates of BOMCC metabolism appeared linear up to 100 pmoles/mL with a Km value determined as 191± 55 μM . Assay conditions of 37 pmoles/mL of T . cruzi CYP51 in bactosomes and 100 μM BOMCC provided a good dynamic range of metabolism . Metabolic activation of 100 μM BOMCC by T . cruzi CYP51 ( 37 pmoles/mL ) was evaluated in the presence of 2% v/v DMSO , methanol or acetonitrile . Incubations were pre-warmed at 37°C before addition of NADPH regenerating system . Production of CHC was measured at 1 minute intervals at 37°C ( Exc 410 nm , Em 460 nm ) for 10 minutes . The average rate of fluorescence units ( AFU ) per minute was compared to a solvent only control ( Fig 4a ) . There did not appear to be a significant decrease in metabolic activity in the presence of 2% v/v DMSO or 2% v/v methanol . With kinetic parameters established , a microtitre-based assay to measure inhibition of recombinantly expressed T . cruzi CYP51 by test compounds was then implemented . Using a selection of ‘azole’ inhibitors , including those previously shown to competitively bind to CYP51 , and known xenobiotic CYP450 inhibitors ( Table 1 ) , the assay was validated to ensure robustness . The incubation times and protein concentrations employed were within the linear range for each assay . Incubations containing 37 pmoles/mL of T . cruzi CYP51 in bactosomes , 100 μM BOMCC , 10 , 3 . 3 , 1 . 0 , 0 . 33 , 0 . 1 , 0 . 033 or 0 . 01 μM test compound ( 2% v/v solvent ) in 50 mM potassium phosphate buffer ( pH 7 . 4 ) , were pre-incubated at 37°C for 5 minutes . Upon addition of NADPH regenerating system ( 7 . 8 mg [28 μM] glucose-6-phosphate , 1 . 7 mg [2 . 2 μM] NADP , 6 units/mL glucose-6-phosphate dehydrogenase in 2% w/v sodium bicarbonate buffer ) production of fluorescent metabolite CHC was measured ( Exc 410 nm , Em 460 nm ) at 1 minute intervals over a 10 minute period . Rates of metabolite production per minute were compared to uninhibited controls and plotted against test compound concentration to obtain an IC50 . The assay was then miniaturized to 20 μL and moved from kinetic to single endpoint stopped readout ( 60 minute reaction before quenching with posaconazole to generate a long stable signal ) to increase throughput . Ketoconazole , itraconazole , posaconazole and miconazole all showed potent inhibition of T . cruzi CYP51 activation of substrate BOMCC with IC50 values of 0 . 014 , 0 . 029 , 0 . 048 and 0 . 057 μM , respectively . Indeed , it is likely that these compounds are even more potent than they appear to be as the T . cruzi CYP51 enzyme concentration used in the assay defines a minimum IC50 of approximately 0 . 02 μM . Fluconazole also inhibited activity ( IC50 0 . 88 μM ) although this was at least 20 fold less inhibitory than observed with other azole type compounds . Methimazole did not appear to inhibit activity ( IC50 > 8 μM ) , perhaps as a result of an inability to adequately fill the active site and prevent substrate binding . Compounds which did not appear to inhibit T . cruzi CYP51 in this assay included known CYP450 inhibitors ticlopidine ( CYP2C19 ) , sulphaphenazole ( CYP2C9 ) , sulphamethoxazole ( CYP2C8/9 ) , quinidine ( CYP2D6 ) and furafylline ( CYP1A2 ) . This was expected as , unlike other drug metabolising CYP450s , CYP51 has a very narrow substrate specificity , being limited to endogenous sterols including eburicol and lanosterol [23] . As previously discussed , addition of an equivalent reductase enzyme ( mosquito , A . gambiae ) was required to deliver metabolic activity of T . cruzi CYP51 . Due to the artificial nature of this pairing it was therefore necessary to confirm that any potent inhibition of CHC production was not the result of indirect inhibition of the electron transfer by NADPH Reductase . Thus , cytochrome c reductase activity of the bactosome preparation was monitored in the presence of all test inhibitor compounds individually . Incubations containing 0 . 82 pmoles/mL of T . cruzi CYP51 in bactosomes , 50 μM cytochrome c , 10 , 3 . 3 , 1 . 0 , 0 . 33 , 0 . 1 , 0 . 033 or 0 . 01 μM test compound ( 2% solvent ) in 50mM potassium phosphate buffer ( pH 7 . 4 ) , were pre-incubated at 37°C for 5 minutes . Absorbance at 550 nm was then measured for 3 minutes to ascertain a background level . After addition of NADPH ( final concentration 80 μg/mL ) , absorbance was further measured over 5 minutes . The rate of reduction of cytochrome c at each test compound concentration was compared to an uninhibited control ( Fig 4b ) . The assay was then miniaturized to 50 μL and moved from kinetic to single endpoint readout ( 60 minutes reaction ) to increase throughput . No decrease in activity of cytochrome c reductase was observed for any of the test inhibitors . Furthermore , subsequent evaluation of a much larger set of T . cruzi CYP51 inhibitors has since been carried out and none have yet decreased cytochrome c reductase activity confirming that inhibition of BOMCC O-dealkylation observed is by direct inhibition of T . cruzi CYP51 and not by its effect on the cytochrome c reductase nor the regeneration system . This strongly suggests that the artificial nature of the pairing of T . cruzi CYP51 with A . gambiae NADPH reductase in the bactosomes used for this assay will not deliver false positives . Identifying the fluorogenic probe BOMCC has enabled development of a fluorescence based high throughput functional assay to determine T . cruzi CYP51 inhibition , suitable for compound triaging of T . cruzi phenotypic screening hits . This assay is now embedded in our Chagas disease screening cascade to prioritise compound series for progression into hit to lead and lead optimisation . Following a phenotypic screen of approximately 200 , 000 compounds against the T . cruzi parasite , inhibition of T . cruzi CYP51 was measured for compounds from 22 trypanocidal series and 16 singleton compounds . Compounds were derived from a variety of sources including diversity and target-centric libraries from the Dundee Drug Discovery Unit and GSK corporate collections . At least three compounds from each chemical series displaying as wide a range of T . cruzi activity as possible were tested ( a total of 129 compounds; Fig 5a ) . With the assumption that those series which show a good correlation between T . cruzi activity and T . cruzi CYP51 enzyme inhibition have a T . cruzi CYP51 mediated mode of action , a remarkable enrichment for the T . cruzi CYP51 mode of action was observed . Correlation between T . cruzi activity and T . cruzi CYP51 inhibition was evident for 11 out of the 22 hit series , with an additional two series showing sporadic association with T . cruzi CYP51 inhibition . Similarly high rates were observed with the singletons , with 11 out of 16 showing antiparasitic activity in line with T . cruzi CYP51 inhibition . 13/22 series and 11/16 singletons were therefore de-prioritised from further medicinal chemistry resourcing and focus given to those hits demonstrating at least one log unit divergence between T . cruzi activity and T . cruzi CYP51 activity , providing confidence that the predominant mode of action was not T . cruzi CYP51 driven . It is clearly important to look at the overall correlation between inhibition of T . cruzi CYP51 and antiparasitic activity for a given series to ensure that they are not tracking . As it transpires , the majority of the chemical series removed from progression bear aza-heterocyclic groups which are already well known to exert inhibition of cytochromes through co-ordination to the haem iron ( Fig 6 ) [9] and could have perhaps been eliminated “by- eye” . However , the great benefit of this assay to our drug discovery effort is that it has allowed identification of T . cruzi active series which do contain potential cytochrome binding motifs , but which do not inhibit T . cruzi CYP51 . For instance , two particularly promising series ( Fig 5b ) having potent T . cruzi activity were found not to have an obvious correlation with T . cruzi CYP51 inhibition in spite of containing heterocycles with known potential for haem-binding in cytochromes . This was in stark contrast to a third series ( Fig 5c ) where T . cruzi activity tracked with T . cruzi CYP51 inhibition , akin to the known human CYP51 training set . The work described here has therefore enabled us to advance the first two series into hit-to-lead studies with a greater degree of confidence and oral activity with compounds from both series has subsequently been demonstrated in mouse models of Chagas disease . The 96 well plate based assay used in this analysis has now been further miniaturized and adapted to automated liquid dispensers and microplate readers for high-throughput screening in 384 well endpoint format . The T . cruzi CYP51 FLINT assay ( Fig 7a ) can be run in 20 μL and the cytochrome c reductase absorbance assay ( Fig 7b ) in 50 μL final volumes . This allows a throughput increase to up to more than 10 , 000 wells in a single experiment with Z’ > 0 . 7 and with a significant reduction in screening cost and time . Sensitivity to described inhibitors was maintained during assay miniaturization .
|
Chagas disease , caused by the parasite Trypanosoma cruzi ( T . cruzi ) , is endemic in Latin America and emerging in North America and Europe through human migration . It is a severe global health problem with 8–10 million people infected and an estimated 12 , 000 deaths annually . Current treatment options are poorly efficacious and have severe side effects . New drugs are therefore urgently required . Two of these potential new drugs , posaconazole and ravuconazole , both targeting an enzyme in T . cruzi called CYP51 , have recently failed in clinical development . Therefore , in light of these recent clinical failures and in order to better balance the overall risk in the drug discovery portfolio for Chagas disease , it has become prudent to assess whether new chemical start points for drug discovery programmes have a mode of action predominantly driven by T . cruzi CYP51 inhibition . In this paper we report a fluorescence based assay to determine whether compounds inhibit T . cruzi CYP51 . This provides a high throughput screen to help prioritise medicinal chemistry resource on those T . cruzi active new chemical series that do not have a mode of action predominantly driven by CYP51 inhibition .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"and",
"Discussion"
] |
[] |
2015
|
Development of a Fluorescence-based Trypanosoma cruzi CYP51 Inhibition Assay for Effective Compound Triaging in Drug Discovery Programmes for Chagas Disease
|
Bangladesh is one of the endemic countries for Visceral Leishmaniasis ( VL ) . Médecins Sans Frontières ( MSF ) ran a VL treatment clinic in the most endemic district ( Fulbaria ) between 2010 and 2013 using a semi-ambulatory regimen for primary VL of 15mg/kg Liposomal Amphotericin-B ( AmBisome ) in three equal doses of 5mg/kg . The main objective of this study was to analyze the effectiveness and safety of this regimen after a 12 month follow-up period by retrospective analysis of routinely collected program data . A secondary objective was to explore risk factors for relapse . Our analysis included 1521 patients who were initially cured , of whom 1278 ( 84% ) and 1179 ( 77 . 5% ) were followed-up at 6 and 12 months , respectively . Cure rates at 6 and 12 months were 98 . 7% ( 1262/1278 ) and 96 . 4% ( 1137/1179 ) , respectively . Most relapses ( 26/39 ) occurred between 6 and 12 months after treatment . Serious adverse events ( SAE ) were recorded for 7 patients ( 0 . 5% ) . Odds of relapse at 12 months were highest in the youngest and oldest age groups . There was some evidence that spleen size measured on discharge ( one month after initiation of treatment ) was associated with risk of relapse: OR=1 . 25 ( 95% CI 1 . 01 to 1 . 55 ) per cm below lower costal margin ( P=0 . 04 ) . Our study demonstrates that 15mg/kg AmBisome in three doses of 5mg/kg is an effective ( >95% cure rate ) and safe ( <1% SAE ) treatment for primary VL in Bangladesh . The majority of relapses occurred between 6 and 12 months , justifying the use of a longer follow-up period when feasible . Assessment of risk of relapse based on easily measured clinical parameters such as spleen size could be incorporated in VL treatment protocols in resource-poor settings where test-of-cure is not always feasible .
Visceral Leishmaniasis ( VL ) , also known as Kala Azar , is a vector borne disease caused by parasites of the genus Leishmania , L . donovani—L . infantum complex , which are transmitted through the bite of an infected sand fly ( mainly genus Phlebotomus , old world , and Lutzomya , new world ) , Leishmaniasis infection in humans presents as cutaneous , muco-cutaneous and visceral . The visceral form is fatal if untreated . VL progressively affects the immune system of the patient , and opportunistic infections are frequently the final cause of death [1 , 2] . VL is endemic in around 80 countries , with over 90% of cases found in India , Bangladesh , Sudan , South Sudan , Ethiopia and Brazil [3] . The Fulbaria district of Bangladesh reported an average annual VL incidence rate of 17 . 8 per 10 , 000 people between 2008 and 2013 [4] . This far exceeds the government’s goal of one reported case per 10 , 000 habitants by the year 2015 , as set in the Bangladesh VL Elimination Program in 2005 [4] . Until 2010 , VL was treated in Bangladesh with Sodium Stibogluconate ( SSG ) or Miltefosine . SSG is given by intramuscular ( IM ) injections daily for 28 days . This course of treatment is painful , toxic , cumbersome for patients , and costly for the health system . In 2008 , the national protocol for VL treatment in Bangladesh shifted to first line treatment with oral Miltefosine for 28 days . Miltefosine has gastro-intestinal side effects , is teratogenic ( requiring 5 months of effective contraception for women of child bearing age , continuing for 3 months post-treatment [5] ) , and is prone to poor adherence if not administered under direct observation [6] . In 2013 , AmBisome single dose ( 10mg/kg ) was adopted as first line treatment , and this regimen is currently being rolled out across the country . Médecins Sans Frontières ( MSF ) has run a VL control project in Bangladesh since 2010 by agreement with the Ministry of Health ( MoH ) . In May 2010 , MSF introduced an innovative first line therapy for primary VL comprising 15mg/kg liposomal amphotericin B ( AmBisome ) intravenous infusion , divided into three doses of 5 mg/kg given over a 5 day period , with only one night of hospitalization . Such a regimen had shown good efficacy and safety in a small clinical trial in India [7] . MSF introduced this regimen because the Bangladesh national VL guideline [8] allowed its use as second line treatment , although its effectiveness and safety had not been evaluated in routine clinical practice in south Asia . It was considered at that time that introduction of other regimens , such as a single dose [9][10] , would require further evidence from clinical trials . The aim of this study was to explore the effectiveness and safety of a short course AmBisome regimen of 15 mg/kg for primary VL under routine program conditions in Bangladesh , and to investigate predictive factors for relapse , with a 12 month follow-up period to assess final outcomes . Clinical studies in VL typically use a follow-up period of 6 months to establish final cure [11] . However , there is some evidence that most relapses occur later than 6 months post-treatment [12] [13] , and wetherefore adopted a 12 month follow-up to allow the assessment of relapse rates up to 12 months post-treatment .
In accordance with the MoH protocol [8] , primary VL was suspected in patients with fever >2 weeks duration , weight loss , splenomegaly or lymphadenopathy , and no history of previous VL . VL was confirmed by rK39 rapid diagnostic test ( RDT ) ( IT-LEISH , Bio-Rad Laboratories , USA ) . The performance of rK39 RDTs in South Asia has been shown to be consistently very high [14–16] . In a setting with a 50% prevalence of VL among clinical suspects this results in a very high positive predictive value . The patient’s general clinical condition , height , weight , spleen size and hemoglobin level ( checked with HemoCue , HemoCue AB , Sweden ) were recorded on admission to the clinic and when the patient was assessed for discharge ( one month after admission ) . Pregnancy testing was done on admission . As the prevalence of malaria in this part of Bangladesh is extremely low , screening for malaria was restricted to vulnerable groups: age <2 or > 60 years and pregnant women , and to suspected malaria cases: patients with acute high fever ( >39°C ) or low hemoglobin ( <8 g/dL ) . Routine HIV testing was not conducted due to the very low prevalence of HIV infection in this population . Patients with confirmed primary VL were treated with intravenous liposomal amphotericin B ( AmBisome , Gilead Pharmaceuticals , Foster City , CA , USA ) using a total dose of 15mg/kg divided into three individual doses of 5 mg/kg: at admission ( day 0 ) ; 24 hours after this dose , and five days after the first dose . Patients were routinely hospitalized for one day , or longer if severely ill . Patients showing a good clinical response ( feeling well , resolution of fever , regression of spleen , and improvement of Hb [11] ) were discharged from treatment after an assessment at 1 month . If there had not been a good clinical response , the possibility of treatment failure was checked by test-of-cure . Test-of-cure was by microscopy of Giemsa-stained smear of splenic aspirate . Patients who had missed any of the three doses were registered as “defaulted” . No parasitological test of cure was done routinely for primary VL cases . Patients presenting with VL relapse were treated with AmBisome ( 3 doses of 15mg/kg if their primary VL had not been treated with AmBisome , 5 doses of 25mg/kg if previously treated with AmBisome ) followed by a test-of-cure after 28 days . Clinical presentation of relapse VL tends to be less severe but , in non-HIV-infected relapse patients , signs and symptoms remain typical ( prolonged undulating fever , splenomegaly , anemia , weight loss ) . Diagnosis of VL relapse was confirmed parasitologically ( by microscopy of Giemsa-stained smear of splenic aspirate ) . All patients were invited to attend a clinical follow-up appointment at 6 and 12 months after treatment or if ill health occurred . At each follow-up appointment a general clinical examination was done , including recording of temperature , spleen size and hemoglobin level . Outcome at each visit was recorded as: clinically cured ( no signs or symptoms of systemic disease ) ; sick but not VL ( patients that were sick at the moment of the follow-up , but did not present signs and symptoms or clinical history suggestive of VL relapse ) ; relapse ( all patients with signs and symptoms of VL , confirmed by spleen aspirate microscopy if there were no contra-indications for splenic aspiration: spleen size >3cm below left costal margin , signs of active bleeding , increased bleeding and clotting time , severe anemia , jaundice , pregnancy , patient unable to remain still ) ; or death . If not attending for follow-up , community health workers actively searched for the patient; if unsuccessful , the patient was classified as "lost to follow-up" . Deaths during the follow-up period were determined by health staff using verbal autopsy to gather information on the cause of death and whether it could be related to VL . Molecular methods were not available for typing Leishmania strains in order to differentiate relapse from re-infection . PKDL was systematically looked for during follow-up , but was not considered a VL treatment failure . For the measurement of effectiveness we classified treatment outcome at one month as ‘initial cure’ for patients whose fever had subsided and who presented with spleen regression , increased hemoglobin level and weight , and improved general condition . Treatment outcomes at 6 and 12 months were categorized as: ‘cured’ ( patients without sign and symptoms of VL ) ; and ‘treatment failure’ ( patients who had symptomatic VL relapse or who had died ) . For the measurement of safety we analyzed the frequencies of reported clinical complications that could be caused by VL treatment ( vomiting , bleeding , and other ) . Severe adverse events ( SAE ) were defined as events leading to suspension or the cessation of VL treatment . Potential determinants of relapse for which there was evidence of an association in univariate analysis were carried forward to a multivariable logistic regression model that was adjusted ( as a priori confounders ) for age and sex . Predictive factors measured at discharge ( one month after initiation of treatment ) were adjusted for their values at baseline ( admission ) and all variables were added one by one to the final model . Continuous variables measured in adults versus children , and on admission versus discharge , were compared using Student’s t test . Data were analyzed with SPSS ( SPSS Inc . Released 2008 . SPSS Statistics for Windows , Version 17 . 0 . Chicago: SPSS Inc . ) and Stata ( StataCorp . 2013 . Stata Statistical Software: Release 13 . College Station , TX: StataCorp LP ) This analysis met the Médecins Sans Frontières International Ethics Review Committee criteria for a study involving the analysis of routinely collected program data . The program utilized a recognized treatment for VL in Bangladesh , and was run in coordination with the Bangladesh Ministry of Health and Family Welfare through a memorandum of understanding , which is the usual procedure for NGOs operating in this context . All electronic data were analyzed anonymously .
The population was young ( 58 . 2% were <18 years old ) and had similar age distributions in both sexes . The male:female ratio was approximately 1:1 across all age groups except among patients age 40+ years ( 3:1 ) . The study population appeared to be under-nourished relative to WHO standards , with a mean BMI of 17 . 5kg/m2 for patients age ≥18 years , and a mean weight-for-height Z-score of −1 . 8 for patients <121 cm tall ( see Table 1 ) . No differences in baseline characteristics were observed between patients from different geographical areas . Adult and pediatric patients presented with similar Hb levels and spleen sizes , and were analyzed as a single patient group . During the treatment period , complications were registered as follows: bleeding 3 . 4% ( 52/1521 ) ; vomiting 9 . 5% ( 144/1521 ) ; and other complications ( including fever , diarrheas , abdominal pain and local rash at the site of the injection ) 0 . 4% ( 6/1521 ) . Serious adverse events ( SAE ) were recorded for 7 patients ( 0 . 5% ) . The remainder had no SAE data , because there was no zero-reporting of adverse events . We cannot analyze the specific details of the SAE because no extra information was recorded in patient files . One of the SAE patients died before attending their first follow-up; of the remainder , none relapsed after 12 months of follow-up . Improvements in key clinical parameters were noted at the one month follow-up: mean Hb levels increased from 8 . 99 to 10 . 87 g/dL ( mean difference 1 . 90 g/dL ( 95% CI 1 . 80 to 1 . 95 ) , P<0 . 001 ) ; mean spleen size decreased from 4 . 6 to 0 . 5 cm ( mean difference −4 . 0 cm ( 95% −3 . 9 to −4 . 2 ) , P<0 . 001 ) ; and mean weight increased from 29 . 7 to 31 . 3 kg ( mean difference 1 . 5 kg ( 95% CI 1 . 4 to 1 . 6 ) , P<0 . 001 ) . Cure rates at 6 and 12 months were 98 . 7% and 96 . 4% , respectively ( Table 2 ) . A ‘best case scenario’ sensitivity analysis , in which we coded all patients lost to follow-up as ‘cured’ ( 16% ( 243/1 , 521 ) at 6 months , 22 . 5% ( 342/1 , 521 ) at 12 months ) , yielded a 6 month cure rate of 98 . 9% and a 12 month cure rate of 97 . 2% . At 6 months follow-up , a total of 13 patients were diagnosed as relapse and 3 deaths were registered . After completing the 12 month follow-up , a total of 39 patients were diagnosed as relapse ( including the previous 13 cases ) : 29 with parasitological confirmation; 10 without ( because splenic aspiration was contra-indicated ) . Of these 10 , none presented with lymphadenopathy . Local staff were not trained to perform bone marrow aspirates hence , lymph or bone marrow aspirates were not taken . No further deaths were registered . Patients followed-up at 12 months did not differ in age , sex and clinical measures on admission and at discharge compared to patients not lost to follow-up , but patients lost to follow-up were more likely to reside in further away sub-districts , which we classified as “distant sub-districts” ( 27 . 5% versus 5 . 9% ) . The three deaths occurred among male patients age >45 years: one entered treatment with a diagnosis of complicated advanced tuberculosis that was probably the main cause of death ( after completing VL treatment he remained as an inpatient and died before the first follow-up ) ; one died between the second and third follow-up from social violence ( according to information provided by family members ) ; and one suffered a severe adverse event ( registered as acute jaundice ) after he completed treatment and died at home before the first follow-up . This was the only case where the death could probably be attributed to VL and/or its treatment . In Table 3 we summarize the results of a logistic regression model with relapse at 12 months as the outcome . We included only those patients with known status at 12 months ( cured or relapsed ) and who had complete data for potential risk factors ( N = 1106 ) . Odds of relapse were highest in the youngest and oldest age groups . There was no association with sex . Anthropometric indices for nutritional status ( BMI in adults , weight-for-height Z-score in children under 121cm ) were not associated with risk of relapse: for adults ( n = 463 ) , OR = 0 . 87 ( 95% CI 0 . 63 to 1 . 20 ) per kg/m2 , P = 0 . 41; for children ( n = 475 ) , OR = 0 . 97 ( 95% CI 0 . 65 to 1 . 44 ) per weight-for-height Z-score , P = 0 . 88 . Larger spleen size at discharge was associated with increased risk of relapse: OR = 1 . 25 ( 95% CI 1 . 01 to 1 . 55 ) per cm below lower costal margin , P = 0 . 04 ) . We found tentative evidence of interaction ( Likelihood ratio test P = 0 . 08 ) between discharge measurements of Hb and spleen size , which suggested an amplified combined effect of low Hb and large spleen size ( S1 Fig ) .
Our study has shown that 15mg/kg AmBisome in three doses of 5mg/kg is an effective ( 96 . 4% cure rate ) and safe ( <1% SAE ) treatment for primary VL in Bangladesh , when judged against internationally accepted parameters for effectiveness ≥95% and safety ( SAE<5% ) for VL treatment [1] . We demonstrated that VL treatment studies in this setting require a follow-up period longer than 6 months if they are to capture the majority of relapses . Whether routine follow-up of discharged patients for this length of time is necessary depends on ease of access to re-treatment for patients who relapse . In settings where access to re-treatment is problematic , routine follow-up may be equally difficult ( and costly ) , but it could help to achieve the VL elimination target ( of <1 case per 10 , 000 people at upazila level in Bangladesh ) set in the Regional Strategic Framework for Elimination of Kala-azar from South East Asia Region [17] . We have also shown that residual splenomegaly is predictive of an increased risk of VL relapse , particularly in conjunction with low levels of hemoglobin . The main strengths of our study are its size ( >1500 patients ) , inclusivity ( 92% of primary VL cases during the period of the study ) , and extended follow-up ( 12-month ) . The main limitation of any cohort study is loss to follow-up . These were low ( 23% at 12 months ) for a study of this design in a resource-poor setting , in part because outreach workers were employed to seek patients who did not attend follow-up appointments . A ‘best case scenario’ analysis showed a maximum cure rate of 97 . 2% . Although we cannot be certain about the validity of this scenario , ease of access to VL treatment in a relatively stable ( non-migratory ) population with multiple means of communication and transport ( and given that apart from distance from facility , we found no differences between patients who were/were not lost to follow-up ) suggest that the true cure rate is unlikely to be lower than 95% . Another limitation was the completeness of SAE reporting because non-occurrence of events was not recorded . Therefore we cannot verify that blank SAE values truly correspond to “treatment complete without SAE” . Although not all milder adverse events will have been reported and we are confident that no severe adverse events were missed , this is a limitation which we would seek to address in future studies by implementing a zero-reporting protocol for SAE . AmBisome has been used to treat VL for more than 10 years [18] , on the basis of evidence from clinical trials . However , evidence for the safety and effectiveness of AmBisome-based regimens under routine program conditions in endemic resource-poor areas remains sparse [19] . This has mainly been due to the high cost of the drug and the neglected status of the disease , factors which MSF has worked to address for many years [20] . In the meantime , studies based on routine data collection from VL treatment programs are an important source of evidence . Our 6-month cure rate ( 98 . 7% ) is consistent with rates reported from other studies of AmBisome [21] , including several conducted in Bihar , India ( a neighboring endemic region ) : Sinha et al reported 98 . 8% with 20mg/Kg in 4 doses over 10 days [22]; Sundar et al reported 95 . 7% with 10mg/Kg in single doses [9] , 98 . 4% with a single 15mg/Kg dose [23] , and 97 . 5% with combined doses of Liposomal Amphotericin B and Miltefosine [24] . In Bangladesh , in the neighboring sub-district of Muktagacha , Mondal et al reported a cure rate of 98% with a single dose 10mg/kg [25] . Our study provides further evidence that the standard 6 month follow-up period for VL studies or treatment programs [1] is insufficient . Several other studies have shown that most relapses occur after 6 months [13] [12] [26] , and we would recommend that clinical trials and treatment programs adopt a 12-month follow-up period if relapse rates are to be measured accurately . However , even this extended length of follow-up is arbitrary , and we cannot discount the possibility that some patients relapsed after 12 months [13] . VL is associated with progressive weight loss and poor nutritional status . Conversely , malnutrition results in impaired immunity and is an important risk factor for severity of clinical VL [27] [28] . The very low anthropometric indices which we found in our patients are consistent with other VL patient cohorts in this population [25] . Whether these low indices are entirely a consequence of VL or whether they also indicate that chronic under-nutrition is a factor in perpetuating the endemicity of VL warrants further research . That spleen size on discharge is a risk factor for VL relapse is consistent with findings from India [13] and South Sudan [29] . We also found tentative evidence of an amplified combined effect of low Hb and large spleen size , which is entirely plausible given that low Hb can be indicative of severity of VL disease ( effect on bone marrow ) , and a massive spleen traps red cells and further depresses Hb [30–32] . Assessment of risk of relapse based on these easily-measured clinical parameters could be incorporated in VL treatment protocols in resource-poor settings where test-of-cure procedures cannot be routinely implemented , with closer and/or extended monitoring of at-risk patients .
Our study has shown that 15mg/kg AmBisome in three doses of 5mg/kg is a safe and effective treatment for primary VL in Bangladesh , and could be an alternative to the current first line regimen of single dose 10mg/kg AmBisome . We have also shown that a follow-up period of 12 months is required to capture the majority of VL relapse cases , and that VL relapse is predicted by low Hb and large spleen size at the end of treatment . Until more evidence is gathered , we would recommend that follow-up be extended to 12 month for all patients or , where this is not feasible , 12-month follow-up could be targeted at patients who present with large spleen at the end of the treatment .
|
Visceral Leishmaniasis ( VL ) is a parasitic disease which is endemic in more than 80 countries , although 90% of cases occur in India , Bangladesh , Sudan , South Sudan , Ethiopia and Brazil . Most treatments are complex , expensive and require long application periods . AmBisome is one of the newest treatments available , but evidence for its safety and effectiveness under routine program conditions in resource-poor endemic areas remains sparse . Médecins Sans Frontières ( MSF ) ran a VL clinic from 2010 until 2014 in Fulbaria District , Bangladesh . Our retrospective study was based on all available data from this clinic , comprising 1521 patients diagnosed with primary VL who were treated with AmBisome 15mg/kg in three equal doses of 5mg/kg . We found that this treatment was safe ( less than 1% of patients experienced a severe adverse event ) and effective ( more than 95% of patients were cured with one treatment ) after 12 months . The youngest and oldest patients , and patients with large spleen size at the end of treatment , were more likely to experience a relapse . More than half of the relapses occurred between 6 and 12 months after treatment , therefore we recommend that clinical trials and treatment protocols adopt a minimum 12-month follow-up period .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion",
"Conclusions"
] |
[] |
2015
|
Effectiveness and Safety of Short Course Liposomal Amphotericin B (AmBisome) as First Line Treatment for Visceral Leishmaniasis in Bangladesh
|
To understand whether any human-specific new genes may be associated with human brain functions , we computationally screened the genetic vulnerable factors identified through Genome-Wide Association Studies and linkage analyses of nicotine addiction and found one human-specific de novo protein-coding gene , FLJ33706 ( alternative gene symbol C20orf203 ) . Cross-species analysis revealed interesting evolutionary paths of how this gene had originated from noncoding DNA sequences: insertion of repeat elements especially Alu contributed to the formation of the first coding exon and six standard splice junctions on the branch leading to humans and chimpanzees , and two subsequent substitutions in the human lineage escaped two stop codons and created an open reading frame of 194 amino acids . We experimentally verified FLJ33706's mRNA and protein expression in the brain . Real-Time PCR in multiple tissues demonstrated that FLJ33706 was most abundantly expressed in brain . Human polymorphism data suggested that FLJ33706 encodes a protein under purifying selection . A specifically designed antibody detected its protein expression across human cortex , cerebellum and midbrain . Immunohistochemistry study in normal human brain cortex revealed the localization of FLJ33706 protein in neurons . Elevated expressions of FLJ33706 were detected in Alzheimer's brain samples , suggesting the role of this novel gene in human-specific pathogenesis of Alzheimer's disease . FLJ33706 provided the strongest evidence so far that human-specific de novo genes can have protein-coding potential and differential protein expression , and be involved in human brain functions .
Many mechanisms for the origination of new genes are known , such as tandem gene duplication , retrotransposition , exon shuffling and gene fusion [1]–[5] . By these mechanisms , the origination of new protein coding genes involved “mother” genes that served as blueprints for the new genes . However , recent comparative genomic analysis identified a few “motherless” or de novo genes in fly and yeast [6]–[9] , which originates from non-coding DNA sequences . It is of great interest to ask whether the human genome also encodes such genes which might contribute to unique human phenotype . Recently Toll-Riera et al identified in silico 15 de novo human genes which seem to have emerged after the split of primates and rodents [10] . However whether these de novo genes encode proteins is unclear due to the lack of protein evidence . More recently Knowles and McLysaght identified in silico three human-specific de novo genes supported by peptides from high-throughput mass spectrum data [11] . These studies , although tremendously interesting , are lacking in two aspects . First , there is no solid protein evidence so far for any of the de novo genes identified—high-throughput mass spectrum data alone as protein evidence can have limitations , as commented by Siepel [12] . Second , none of these genes has been linked to human specific phenotype . Could any de novo genes be associated with human unique biology , especially to brain functions ? In our work , we were interested in finding de novo genes associated with nicotine addiction . We took advantage of the recently available high-throughput data from genome-wide association studies ( GWAS ) and data from the more traditional linkage analyses . Unlike candidate gene association studies that usually start with a known gene , GWAS and linkage analyses are hypothesis-free and thus can link previously uncharacterized genes to addiction . Despite the great potentials , current GWAS results are under-analyzed and under-utilized . There is a need for computational protocols to sift through the GWAS results for interesting genes .
Here we carefully re-analyzed results from two published GWAS [13] , [14] and two linkage analyses [15] , [16] for nicotine addiction and looked for genes that ( i ) show statistical significance in both GWAS and both linkage analyses; and ( ii ) have a complete Open Reading Frame ( ORF ) that has no identifiable homologues in other species . We found an interesting gene , FLJ33706 ( alternative gene symbol C20orf203 ) . Both GWAS identified rs17123507 , an SNP located in the 3′UTR of FLJ33706 , as significantly associated with susceptibility to nicotine addiction [13] , [14] . Both linkage studies also implicated this region in ‘heavy-smoking quantitative trait’ in individuals of European ancestry [15] , [16] . These genetics data established the genomic region of FLJ33706 as one of the 10 ‘convergent susceptible points’ for nicotine addiction [16] . However , FLJ33706 was not directly reported as a candidate gene to explain the genetic vulnerabilities in any of the four studies , and to date , FLJ33706 remains an un-studied gene . In the next steps of our work , we demonstrated that FLJ33706 is an interesting human-specific de novo protein-coding gene . We traced how this fascinating gene originated out of noncoding DNA sequence and experimentally studied its population genetics , mRNA expression , protein expression , and cellular localization . FLJ33706 is located on Chromosome 20q11 . 21 . Little is known about this gene: it has no publication , no detectable protein domain by InterPro [17] , and no BLAST hit to any other known protein sequences . Four mRNAs and four spliced ESTs in GenBank map to this locus , supporting the expression of FLJ33706 at the transcription level . The UniProtKB/TrEMBL database provided a computationally translated ORF and label it a “predicted protein” ( Accession Number: B8JHY2_HUMAN ) [18] , but the UCSC genome browser and NCBI Entrez Gene database marked it as a “non-coding RNA” [19] , [20] . We re-sequenced all five available EST clones ( see details in Materials and Methods ) and inferred the gene structure of FLJ33706 ( GenBank Accession Number: GU931820 ) . The whole locus covers a 42 . 3 Kb genomic region , encoding a 5 , 093 bp polyadenylated transcript separated by five standard introns marked with GT-AG splicing junctions . A putative open reading frame ( ORF ) with 194 codons is located in exons 3 and 4 ( Figure 1 ) . The 44-way vertebrate syntenic genome-alignment tracks of the UCSC browser [19] showed that the DNA segment where FLJ33706 gene is located emerged in the eutherian mammals , since it is completely absent from all outgroups ranging from marsupials to lamprey ( Supplementary Figure S1 ) . Although this locus predated the radiation of modern mammals , the full splicing structure appeared at a much later time . Specifically , syntenic alignments flanking five splicing junctions ( Figure 2 ) revealed that non-primate mammals only encode the first standard splicing junction . For the remaining four introns , non-primate mammals used non-standard junctions if they spliced these regions out at all . Most likely the last four introns were not spliced . Furthermore , only hominoid used GT-AG for the third intron , while the possible ancestral states shared by rhesus monkey and mouse lemur armadillo is GA-AG ( Figure 2 ) . Such difference across the splicing junctions indicated that the FLJ33706 locus must have undergone multiple-step changes in order to acquire the present relatively complex gene structure in human . Manual inspection of the gene structure and vertebrate genome comparisons showed that newly inserted repeat elements , especially Alu sequences , contributed substantially to the formation of the first coding exon and the six standard splice junctions on the branch leading to the hominoid ( Figure 1 , 2 ) . Specifically , the splicing acceptor of the second intron , the donor and acceptor of the fourth intron , and the splicing donor of the last intron were derived from Alu sequences . In addition , Alu contributed to 71% of the first coding exon and 16% of the total ORF . This finding is consistent with other reports that transposable elements can contribute to the creation of both protein-coding regions and splice junctions [10] , [21] . The putative ORF of FLJ33706 is human-specific . Sequence alignments across multiple primates including human , chimp , gorilla , orangutan , rhesus monkey and marmoset showed that the FLJ33706 ORF emerged only on the human lineage after the divergence of human and chimpanzee by the introduction of five point mutations , including two important mutations that escaped two ancestral frame-disrupting features , TAG—>TGG at amino acid position 28 and GGAA—>G-AA at amino acid position 106 ( Figure 3 ) . Chimpanzee seems to share the ancestral status for both of these sites . This is unlikely to be an artifact caused by sequencing error because the sequencing quality of the chimpanzee genome in this region is quite high . For example , TAG is supported by six chimp reads ( Supplementary Figure S2 ) . Thus , FLJ33706 is likely a bona fide human-specific de novo protein-coding gene . As aforementioned , eight spliced mRNA and EST sequences support the transcription of FLJ33706 . These transcripts were mainly cloned from brain libraries , suggesting brain-enriched expression of FLJ33706 . No mRNA or EST in Genbank from any other species could be reliably mapped to the orthologous genomic locus or to FLJ33706—only one unspliced Sus scrofa EST ( BI343741 ) could be mapped to the first 3′ untranslated region ( UTR ) of FLJ33706 . The GEO [22] microarray database included a databset GSE7094 which profiled five tissues ( cortex , fibroblast , pancreas , testis and thymus ) in rhesus monkey . Re-analysis of the data showed low expression signal in Rhesus Macaque ( normalized expression intensity 2 . 2∼2 . 7 ) . In summary , both EST and microarray data indicated that FLJ33706 has low or non-existent transcription in non-hominoid mammals . We further experimentally quantified FLJ33706 mRNA levels in eight human peripheral tissues and eight human brain regions using the TaqMan technique with FAM-labeled probe hybridized across exon3 and exon4 . We observed that FLJ33706 mRNA was significantly enriched in the brain , especially in regions implicated in cognitive abilities ( Figure 4 , Supplementary Table S1 ) . The mRNA expression levels of FLJ33706 in cortex and hippocampus were comparable to those of the neuronal specific isoform of brain-derived neurotrophic factor ( BDNF1 ) , although lower than those of the calcium activated isoform ( BDNF4 ) [23] . The biased tissue expression patterns of FLJ33706 and the comparable expression levels between FLJ33706 and BDNF1 provided further support that FLJ33706 might be a functional gene . To explore whether or not FLJ33706 may have protein coding potential , we first performed population genetics analysis of 90 individuals including all major sub-populations ( Supplementary Table S2 ) to investigate whether this putative coding region , especially the nonsynonymous sites , was under more constraint . We sequenced the coding region and 1 Kb flanking regions of the FLJ33706 locus in the 90 individuals . No frame-disrupting mutation was found , which suggested some degree of protein-level constraint . Moreover , the nonsynonymous sites showed the strongest constraint ( nucleotide diversity π of 5×10−5 ) ( Table 1 , Supplementary Table S2 ) . By contrast , synonymous sites had an order of magnitude larger π ( 4×10−4 ) . We further tested whether this difference departed from neutral assumptions using Hudson's formula [24] . Despite of the small size of this putative protein , the comparison still yielded a marginally significant p of 0 . 1 , which suggested that the nonsynonymous sites did evolve under more constraint . Finally , the whole coding region had lower nucleotide diversity π compared to its immediate flanking regions , the second intron or the 3′ UTR ( Supplementary Figure S3 ) . In summary , population genetics analysis suggested that FLJ33706 potentially encoded a protein under purifying selection . However , protein-coding potential of FLJ33706 suggested by population genetics analysis was still not conclusive . To explore whether or not FLJ33706 actually encodes the 194-codon protein , we developed FLJ33706-specific antibody and performed Western blot analyses . We designed a 17-amino-acid antigenic peptide , CTSKAQRVHPQPSHQRQ , corresponding to the non-repetitive region ( residues 68–83 ) of the FLJ33706 putative protein plus a cystine at the N-terminus to facilitate conjugation to an adjuvant . The epitope sequence had no homology with the coding peptides of Alu or other repeat elements and could not match any other proteins in NCBI NR database [20] . This peptide was synthesized and used to immunize rabbits . The FLJ33706-specific anti-serum was produced from a responsive animal after initial and boosting immunizations . Using this anti-serum as the primary antibody , Western blot assay detected a band with apparent molecular mass of 22 kDa , which was consistent with the predicted molecular weight of the FLJ33706-encoded protein , in human brain cortex ( Figure 5A ) . This band was not present when pre-immune serum was used or when the antibody was pre-absorbed with excess synthetic FLJ33706 antigenic peptides ( Figure 5A ) [25] . We further expressed FLJ33706 recombination protein with His-Tag in E . coli expression strain to evaluate the specificity of FLJ33706 antibody in Western blot assays . As expected , the band with apparent molecular mass of 22 kDa was detected in transformed E . coli samples by both His-Tag specific antibody and the aforementioned FLJ33706 antibody , but not in wild-type E . coli samples ( Figure 5B ) . These results provided verification of the antibody . Using this verified FLJ33706-specific antibody , we studied the expression and localization of FLJ33706 . We first identified the expression of FLJ33706 in three human brain regions: cortex , cerebellum and midbrain . The specific band could be detected in all human samples but not mouse samples as negative controls ( Figure 5C ) . We then performed within-species studies using cortex samples from seven different human brains and observed FLJ33706 expression in all samples , with some variation in protein expression levels ( Figure 5D ) . We further performed immunohistochemistry studies of FLJ33706 by high-resolution confocal imaging in normal human cortex slides stained with beta-tubulin-III . The clear co-localization signals indicated cellular localization of FLJ33706 protein in human neurons ( Figure 6 ) . Could FLJ33706 be involved in other human brain-related pathogenesis such as AD ? As a preliminary study , we measured the transcriptional level of FLJ33706 in the middle fontal gyrus ( Brodmann area 46 ) of 20 AD brains and 18 normal brains using the TaqMan-based Real-Time PCR system . The expression level of FLJ33706 in AD brains was significantly elevated ( Mann Whitney Test , p = 0 . 027 ) ( Supplementary Figure S4 ) . This finding implicated FLJ33706 as a potential candidate gene for studying the human-specific pathogenesis underlying Alzheimer's disease [26] .
In previous works , only one of the de novo genes in yeast and three in human had some high-throughput mass spectrum evidence of protein coding potential [7] , [11] . However high-throughput mass spectrum data can be noisy and peptide identification is dependent on the algorithms and search parameters . Our results on FLJ33706 provided the strongest experimental evidence so far of protein expression and differential protein expression of a de novo gene . We experimentally verified the existence of the predicted ORF in human , and observed two frame-disrupting features in chimpanzee that would prevent this ORF from being translated . Moreover , these two features are shared by multiple non-human primates , which suggest that this ORF did not exist in the ancestral status . Identification of ancestral frame-disrupting features is a common strategy to identify species-specific de novo proteins [27] , [28] . Ideally , we would want to use chimpanzee tissues as negative controls in the Western blot assays . Unfortunately , it proved impossible for us to obtain chimpanzee postmortem samples , especially brain regions , due to our limited resources . Despite this , all our current evidence supports FLJ33706 as a human-specific de novo protein . The recently published genome-wide scan by Knowles and McLysaght identified three human-specific de novo protein-coding genes [11] but failed to identify FLJ33706 . The authors used the automated annotations by Ensembl ( version 47 ) which incorrectly annotated FLJ33706 as having an orthologous protein-coding gene in chimpanzee ( ENSPTRG00000030588 ) . However , as we described before , the chimpanzee locus consists of two frame-disrupting features . In order to make an intact ORF , Ensembl's automatic annotation pipeline made these two features ( “TAG” and “G” ) as extra tiny introns inside the frame . Such events are extremely unlikely because very few human introns are smaller than 80 bps [17] . In other words , misannotation of Ensembl have likely resulted in the failure of Knowles and McLysaght [11] to discover FLJ33706 . Siepel commented on the importance of distinguishing true de novo genes from genes that were functional in ancestral genomes but lost in multiple lineages [12] . In the case of FLJ33706 , the latter scenario is highly unlikely . First , we traced the whole evolutionary history of FLJ33706 across vertebrates and found that only human has an intact ORF . If this gene were functional in ancestral mammals , then there would have to be too many independent gene loss events , which is highly unlikely . Second , parallel loss for the same locus in different lineages requires that this locus be in some sort of mutational hot spot [12] . Our population survey showed that FLJ33706 does not have an unusually high level of polymorphism ( θ ∼0 . 001 which is comparable to the genome-wide background level of 1×10−3 ) [29] . Thus , at least in human , this locus is not generally permissive for mutation . In summary , FLJ33706 is a bona fide de novo gene . Siepel proposed a few features of de novo genes [12]: de novo gene products are usually small with less than 200 amino acids because of the difficulty in de novo gene origination; they are often derived from the antisense strand of a pre-existing gene so that they might be able to re-use the transcriptional context; repeats elements might be involved in origination of some de novo genes as shown for the gene hydra in D . melanogaster [8] . FLJ33706 showed similar features: it encodes a small protein of 194 amino-acids; although it is not derived from the antisense strand of another gene , it is located in a gene-dense region with two other genes in its immediate flanking regions ( <30 kb distance ) and thus the local chromatin structure might be open , which renders transcription more permissive; and finally , the primate-specific repeat element , Alu , contributed to origination of multiple introns and a portion of the coding region . The small protein size and human-specific nature of FLJ33706 resulted in insufficient statistical power for many evolutionary tests . Nevertheless , we were still able to detect that this locus deviates from neutral expectation . Polymorphism distribution across different functional sites including non-synonymous sites , synonymous sites , UTR and introns suggested that FLJ33706 is subject to functional constraint . Base-level conservation score calculated by PhyloP [30] based on placental mammal genome alignment showed that introns 2 and 3 are enriched with fast-evolving nucleotides ( Supplementary Figure S5 ) which suggested that the emergence of these two introns in primate might be driven by positive selection . Although this locus existed since at least 80 million years ago ( the time for mammalian radiation ) , its complete splicing structure encoding five standard splicing junctions is younger than 38 million years ( human and rhesus monkey divergence time ) [www . genome . gov/Pages/Research/Sequencing/SeqProposals/PrimateSEQ012306 . pdf] . It is possible that FLJ33706 is already transcribed in the hominoid ancestor at low abundance . Thus , human FLJ33706 protein may have evolved out of a noncoding RNA which evolved out of noncoding DNA . Furthermore , FLJ33706 are mainly expressed in human brain , with more than two folds higher expression in cortex compared to testis . By contrast , its ortholog in rhesus monkey seems to have low expression intensity in major tissues and non-differential abundance between cortex and testis . Thus , FLJ33706 not only acquired more complicated gene structure , but refined its expression profile in the human lineage . As mentioned above , an addiction-linked SNP rs17123507 is located in the gene region of FLJ33706 , confirmed by two GWAS and two linkage analyses . To clarify whether this SNP is the ‘causative’ SNP of addiction susceptibility within its haplotype block , we used HapMap data to identify all SNPs showed strong linkage disequilibrium ( r2≥0 . 8 ) with rs17123507 . rs17123507 was the only one located in the exon region ( 3′UTR ) of FLJ33706 among a tandem set of putative binding sites of let-7 , a brain-expressed miRNA implicated in neuron specification [31] . All other SNPs were located in intronic or intergenic regions without any annotations or detectable signals of regulatory elements . Thus , rs17123507 was the most possible ‘causative’ SNP within the haplotype block that convey addiction susceptibility . We also found that FLJ33706 expressions were up-regulated in AD brains . Thus FLJ33706 is likely involved in a range of human brain functions and pathogenesis . However , exactly how FLJ33706 affects human brain functions and exactly why both addiction and AD might be implicated remain unknown and are interesting questions for future studies . GWAS provides invaluable links between genes and diseases/phenotypes at high throughput . During the past few years , GWAS have identified numerous genetic variations that contribute to susceptibilities underlying various complex diseases . However , GWAS data is often under-analyzed and poorly interpreted . Our work provides a computational protocol for identifying and studying interesting candidate genes from GWAS of not only addiction , but also other diseases and phenotypes . On the other hand , the studies of the functions of novel genes are time-consuming and often involve much guesswork . Our work demonstrated the feasibility of integrating the rapidly accumulating data from GWAS and linkage analyses to associate novel genes with human diseases and phenotypes . Our work is a good example of how computational screening of existing biological data can lead to interesting , experimentally verifiable discoveries . Although we spent much effort to experimentally verify the gene and protein expression of FLJ33706 , the most novel part of our contribution is in fact how we had computationally selected this hidden gem from the human genome in the first place . More specifically , our work can serve as a model for future studies of de novo species-specific protein-coding genes that would start from computational and evolutionary analyses similar to what we have done here . In conclusion , our data provided the strongest evidence so far for a human-specific de novo protein and its association with human brain functions . It had been well accepted that protein amino acid changes , protein family expansion and shrinkage , and cis-regulatory element changes contributed to human brain evolution [32] . Our study suggested that motherless new genes may be an under-appreciated source of new brain functions .
This study was conducted according to the principles expressed in the Declaration of Helsinki . Human tissues were obtained from Department of Pathology , Johns Hopkins Medical School and the NICHD Brain and Tissue Bank , which have been approved by the Institutional Review Board of Johns Hopkins Medical School and University of Maryland , Baltimore , Maryland , USA . All animals were handled in strict accordance with good animal practice as defined by the relevant national and/or local animal welfare bodies , and all animal work was approved . Brain tissues from 20 Alzheimer's disease ( AD ) patients and 18 non-AD control individuals were obtained post mortem ( Department of Pathology , Johns Hopkins Medical Institutions ) . For each individual sample , a portion of medial frontal gyrus ( Brodmann area 46 ) was prepared for extraction of total RNA . Frontal cortex , midbrain , and cerebellum brain regions were obtained from the NICHD Brain and Tissue Bank for Developmental Disorders at the University of Maryland . Human brain samples used in immunohistochemistry studies were ordered from the Folio Company . The human DNA samples from 90 different individuals were order from Coriell Cell Repositories . Mouse brain samples were prepared in accordance with previous studies [25] , [33] . Available EST clones for FLJ33706 ( Entrez GeneID: 284805 ) , including BC105014 , BG820670 , AW196294 , H08894 and AI301139 , were purchased from Invitrogen CloneRanger™ and sequenced by Invitrogen . Exons of FLJ33706 were then assembled with Sequencher software ( Gene Codes Corporation , USA ) using publicly available reliable mRNAs , spliced ESTs and results from our re-sequenced clones . RNA isolation , cDNA synthesis , and real-time PCR were performed as described previously [25] , using glyceraldehyde-3-phosphate dehydrogenase ( Applied Biosystems ) as an endogenous control . Brain region and peripheral tissue RNAs were purchased from Clonetech . FLJ33706 specific Fam-labelled MGB probe across exon 3 and 4 ( 5′-TGA GCC GGG CCA CAT-3′ ) and PCR primers ( Forward: 5′-TCC CTT TAC AAA AAC TGG AAT GC-3′; and Reverse: 5′-GCA GTG AGT CCA GCC AAG ACT-3′ ) were designed to detect the transcript . Relative quantity was calculated using expression means of human leukocyte as references . Expression levels of two BDNF isoforms in human cortex were used as references , following the protocols proposed in Liu et al [23] , [25] . In order to test the functional constraint of the putative small protein encoded by FLJ33706 , we sequenced 90 human individuals in different populations ( Supplementary Table S2 ) . DNA samples were purchased from the Coriell Institute for Medical Research . The FLJ33706 locus including the coding sequence and 1 Kb flanking regions ( intron or untranslated regions ) were PCR-amplified using primers designed by Oligo ( http://www . oligo . net ) . When necessary , we ran multiple PCR experiments to amplify the full-length region . PCR bands were sent to Invitrogen for sequencing . For each copy , four walking reactions were performed . Subsequently , we used Phred , Phrap and Consed [34] , [35] to assemble the FLJ33706 locus for each individual . Single nucleotide polymorphisms ( SNPs ) were identified with Polyphred [36] and Polyscan [37] . Specifically , homozygous or heterozygous SNPs were called by Polyphred first . We retained those highly reliable SNPs with Polyphred score of 99 . For SNPs with a score lower than 99 , we retained them only if they were also identified by Polyscan . We used DnaSP v4 . 50 [38] to calculate the statistics of polymorphisms . We calculated the probability of the number of observed segregation sites in nonsynonymous sites on a hypothetical θ ( e . g . the one in synonymous sites ) by following the recursive equations [24]: Where , l , n and s are defined as the length of region of interest , the number of alleles and the number of segregation sites , respectively . Qn ( i ) indicates the probability that i mutations occur when there are n ancestral lineages , while Pn ( s ) indicates the probability that s sites segregate in a sample of n individuals . We found in Affymetrix Rhesus Macaque Genome Array a probeset MmugDNA . 22336 . 1 . S1 for the orthologous locus of FLJ33706 . We also found a GEO [22] dataset , GSE7094 , which profiled five tissues ( cortex , fibroblast , pancreas , testis and thymus ) in a rhesus monkey with six replicates for each sample [39] . We downloaded GSE7094 raw array files from NCBI GEO database [22] . We used R and Bioconductor [40] platform to handle this data . Specifically , we used GCRMA [41] to do background subtraction , normalization and probe summarization , and Microarray Suite , version 5 . 0 ( MAS5; Affymetrix ) to call presence or absence . We expressed FLJ33706 recombination protein in E . coli expression strain . The full-length coding region of FLJ33706 was obtained by PCR amplification using an isolated human genomic library as the template . The PCR products were ligated by T4 DNA ligase and the resulting full-length fragment was sub cloned into the pET-28a expression vector with Poly His tag . The resulting recombinant plasmids were verified by DNA sequencing , followed by transformation into the E . coli expression strain BL21 ( DE3 ) . E . coli samples before and after the transformation were prepared for Western blot assays . A 17-amino-acid peptide with sequence CTSKAQRVHPQPSHQRQ that corresponded to the unique residuals 68–83 of FLJ33706 putative protein was synthesized ( cystine was added to conjugate to keyhole limpet hemocyanin ) and used to immunize rabbits ( Genemed Synthesis , Inc . , San Antonia , TX , USA ) . The peptide sequence is highly antigenic and lacks detectable homologues in any mammalian genomes based on BLASTP . The FLJ33706-specific anti-serum was produced in a favourable animal after initial and boosting immunizations . Protein levels were quantitated using Bradford assays and 50 µg protein aliquots of supernatant were electrophoresed using 10% SDS-polyacrylamide gels and Western blot analysis was performed as described previously [25] . FLJ33706 anti-serum that was diluted 1∶5000 and the pre-immune serum that was diluted with 1∶5000 were used to replace anti-FLJ33706 serum . The synthetic peptide ( 100 µg/ml ) was incubated with primary antiserum that had been pre-absorbed 2 h at room temperature for the competition assay [25] . Western blot assays with E . coli expressed FLJ33706 recombination protein ( with His-Tag ) were also introduced to evaluate the specificity of FLJ33706 antibody , in which anti-FLJ33706 and anti-His tag was diluted at 1∶5000 and 1∶500 , respectively . Immunohistochemistry study of FLJ33706 in human brain cortex was performed as previously described [42] . Antiserum of FLJ33706 is produced as mentioned above ( 1∶400 ) , and antibody against beta-tubulin III was ordered from Sigma ( 1∶200 ) .
|
For decades , gene duplication , retrotranspositions and gene fusions were believed to be major ways to increase gene number . All involve “mother” genes as the “building blocks” for new genes . However , several recently identified “motherless” genes challenged the idea in that some proteins might have emerged de novo from ancestral non-coding DNAs . Did any such genes emerge in human after the divergence from chimpanzee ? If yes , such genes might help understand what makes us human . Here we report the first experimentally verified case of a human-specific protein-coding gene , FLJ33706 ( alternative gene symbol C20orf203 ) , that originated de novo since the divergence of human and chimpanzee . FLJ33706 was formed by the insertion of repeat elements , especially Alu sequences , that contributed to the formation of the first coding exon and six standard splice junctions , followed by two human-specific substitutions that escaped stop codons . The functional protein-coding features of the FLJ33706 gene are supported by population genetics , transcriptome profiling , Western-blot and immunohistochemistry assays . Data suggest that FLJ33706 may be involved in nicotine addiction and Alzheimer's disease . FLJ33706 provided the strongest evidence so far that human-specific de novo genes can have protein-coding potential and be involved in human brain functions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"evolutionary",
"biology/human",
"evolution",
"evolutionary",
"biology/evolutionary",
"and",
"comparative",
"genetics"
] |
2010
|
A Human-Specific De Novo Protein-Coding Gene Associated with Human Brain Functions
|
Poxviruses contain large dsDNA genomes encoding numerous open reading frames that manipulate cellular signalling pathways and interfere with the host immune response . The NF-κB signalling cascade is an important mediator of innate immunity and inflammation , and is tightly regulated by ubiquitination at several key points . A critical step in NF-κB activation is the ubiquitination and degradation of the inhibitor of kappaB ( IκBα ) , by the cellular SCFβ-TRCP ubiquitin ligase complex . We show here that upon stimulation with TNFα or IL-1β , Orthopoxvirus-infected cells displayed an accumulation of phosphorylated IκBα , indicating that NF-κB activation was inhibited during poxvirus infection . Ectromelia virus is the causative agent of lethal mousepox , a natural disease that is fatal in mice . Previously , we identified a family of four ectromelia virus genes ( EVM002 , EVM005 , EVM154 and EVM165 ) that contain N-terminal ankyrin repeats and C-terminal F-box domains that interact with the cellular SCF ubiquitin ligase complex . Since degradation of IκBα is catalyzed by the SCFβ-TRCP ubiquitin ligase , we investigated the role of the ectromelia virus ankyrin/F-box protein , EVM005 , in the regulation of NF-κB . Expression of Flag-EVM005 inhibited both TNFα- and IL-1β-stimulated IκBα degradation and p65 nuclear translocation . Inhibition of the NF-κB pathway by EVM005 was dependent on the F-box domain , and interaction with the SCF complex . Additionally , ectromelia virus devoid of EVM005 was shown to inhibit NF-κB activation , despite lacking the EVM005 open reading frame . Finally , ectromelia virus devoid of EVM005 was attenuated in both A/NCR and C57BL/6 mouse models , indicating that EVM005 is required for virulence and immune regulation in vivo .
The NF-κB family of transcription factors activate potent pro-inflammatory and anti-viral immune responses that are activated by a variety of signalling pathways [1] , [2] . The family consists of five members , p50 , p52 , p65 ( RelA ) , RelB , and c-Rel , which function as homo- or heterodimers to activate specific genes . The best-characterized NF-κB dimer is the p50/p65 heterodimer , which is held inactive in the cytoplasm by the inhibitor of κB ( IκBα ) [1] , [2] . Signalling cascades initiated by both tumour necrosis factor α ( TNFα ) and interleukin 1β ( IL-1β ) trigger the activation of a set of kinases known as the IκB kinase ( IKK ) complex , which is composed of IKKα , IKKβ and IKKγ/NF-κB essential modifier ( NEMO ) [2] . Upon activation of the IKK complex , IKKβ phosphorylates IκBα on serines 32 and 36 , targeting IκBα for polyubiquitination and degradation by the 26S proteasome [1] , [2] . The SCF ( Skp1/Cul1/F-box ) ubiquitin ligase recruits phospho-IκBα through the F-box domain-containing adaptor protein , β-TRCP , resulting in the degradation of IκBα , and translocation of the p50/p65 heterodimer into the nucleus [1] , [2] . Regulation of NF-κB signalling is common amongst most viruses , with each virus employing a combination of specifically tailored strategies [3]–[5] . For example , human immunodeficiency virus ( HIV ) , human T-lymphotropic virus type 1 ( HTLV-1 ) , hepatitis B virus ( HBV ) , and Epstein-Barr virus ( EBV ) activate the NF-κB signalling pathway [4] . Virus activation of the NF-κB pathway could serve several roles . For instance , viruses that lack anti-apoptotic mechanisms may activate NF-κB to prolong the life of the infected cell in order to complete the viral replication cycle . In the case of EBV , constitutive activation of NF-κB leads to the up-regulation of NF-κB-regulated pro-survival proteins during latency [6] . Alternatively , HIV-1 contains NF-κB binding sites in the long terminal repeat ( LTR ) region of the genome that mediate HIV-1 gene expression [7] . In contrast , other viruses encode proteins that specifically inhibit NF-κB signalling [3]–[5] . For example , the V and C proteins encoded by the Paramyxoviridae associate with the STAT family of transcription factors , thus inhibiting the interferon response and NF-κB activation [8] . Moreover , African Swine Fever Virus encodes a homolog to IκBα that sequesters p65 in the cytoplasm following IκBα degradation [9] . Overall , the varieties of viral proteins that manipulate NF-κB indicate the importance of the long and varied relationship with NF-κB . The inhibition of NF-κB by poxviruses has become a growing area of interest [5] . The Poxviridae is composed of viruses possessing large dsDNA genomes , encoding between 150 to 300 open reading frames [10] . Poxviruses are unique amongst DNA viruses in that they replicate in the cytoplasm , within DNA-rich regions termed “virus factories” [10] . Members of the Orthopoxvirus genus are well studied , and include variola virus , vaccinia virus ( VACV ) , monkeypox virus , cowpox virus ( CPXV ) , and the mouse-specific pathogen , ectromelia virus ( ECTV ) [11] . Poxviruses are renowned for the plethora of immune evasion mechanisms they encode; including mechanisms that regulate NF-κB [5] , [12] , [13] . One of the first identified mediators of NF-κB activation was M-T2 , a secreted soluble virus version of the tumor necrosis factor receptor ( vTNFR ) [14] , [15] . Soluble vTNFRs and vIL-1Rs were subsequently identified in a variety of poxviruses [13] . More recently , focus has shifted to the identification of intracellular inhibitors of NF-κB encoded by poxviruses [5] . VACV encodes three proteins , K7 , A46 , and A52 , which contain Toll/IL-1 receptor ( TIR ) cytoplasmic domains and disrupt NF-κB activation triggered through the IL-1/Toll receptor pathway [16]–[18] . Additionally , the VACV-encoded proteins , B14 , M2 , K1 , A49 , and N1 , disrupt NF-κB activation triggered through the TNFR pathway [19]–[22] . These proteins function at different points in the signalling cascade , clearly highlighting the importance of NF-κB inhibition during infection [19]–[24] . Recently , we identified a family of four ankyrin/F-box proteins encoded by ECTV , EVM002 , EVM005 , EVM154 ( recently renamed EVM159 ) , and EVM165 ( recently renamed EVM170 ) that interact with the cellular SCF ubiquitin ligase complex [25] . The ECTV-encoded proteins contain N-terminal ankyrin repeats in conjunction with a C-terminal F-box domain; similar viral ankyrin/F-box proteins have been found in a wide range of poxviruses [26] . To date , no cellular F-box proteins have been found in conjunction with ankyrin repeats , suggesting that poxviruses , including ECTV , have evolved a novel set of genes to regulate the cellular SCF ubiquitin ligase . Multiple orthologs for EVM002 , EVM154 , and EVM165 exist within the poxvirus family; however , EVM005 has only one ortholog , CPXV-BR011 , in CPXV virus strain Brighton Red , suggesting that EVM005 and CPXV-BR011 may play an important role that is specific to ECTV and CPXV . Since degradation of IκBα is catalyzed by the SCFβ-TRCP ubiquitin ligase , we investigated the role of EVM005 in regulation of IκBα degradation . Here , we show that cells infected with ECTV and stimulated with TNFα or IL-1β accumulate phosphorylated IκBα , indicating that IκBα is stabilized and not degraded . Ectopic expression of Flag-EVM005 inhibited both TNFα- and IL-1β-stimulated IκBα degradation and subsequent nuclear translocation of NF-κB; however , deletion of the EVM005 F-box domain resulted in activation of NF-κB . ECTV devoid of EVM005 , ECTV-Δ005 , inhibited NF-κB activation . Finally , we demonstrated that EVM005 is a critical virulence factor , since ECTV-Δ005 was attenuated in both A/NCR and C57BL/6 mice compared to wild type ECTV . These data demonstrate a novel role for poxvirus-encoded ankyrin/F-box proteins in regulation of the SCF ubiquitin ligase and NF-κB signalling .
The NF-κB signalling cascade activates a family of transcription factors responsible for initiating the pro-inflammatory response and antiviral innate immunity [1] , [2] . Recent evidence indicates that many poxviruses encode proteins that tightly regulate the activation of NF-κB through the expression of secreted and intracellular factors [5] , [12] . Unlike strains of VACV , ECTV lacks M2 , K7 , B14 , A49 and A52 , all of which are important inhibitors of NF-κB activation [16] , [18] , [21] , [23] , [24] . Given the absence of these inhibitors , we sought to determine if ECTV infection inhibited NF-κB activation . Since the degradation of IκBα is crucial for activation of the NF-κB pathway , we examined the kinetics of IκBα degradation during infection . HeLa cells were mock-infected , or infected with ECTV or VACV for 12 hours and treated with TNFα for up to 120 minutes . Mock-infected cells treated with TNFα showed a typical pattern of IκBα degradation kinetics ( Fig . 1A ) . As early as 10 minutes post-TNFα treatment , mock-infected cells showed phosphorylated IκBα , as indicated by a doublet , which was subsequently degraded ( Fig . 1A and B ) [27] , [28] . ECTV- and VACV-infected cells treated with TNFα also showed obvious phosphorylation of IκBα ( Fig . 1A and B ) ; however , the levels of both IκBα and phospho-IκBα were sustained compared to mock-infected cells ( Fig . 1A and B ) . Interestingly , we did observe that lower levels of phospho-IκBα accumulated in cells infected with VACV compared to ECTV , and accumulation was delayed in comparison to ECTV-infected cells ( Fig . 1A ) . Western blotting for I5L , a late poxviral protein , and cellular β-tubulin were used as loading controls ( Fig . 1C and D ) [25] . Similar observations were also seen following treatment with IL-1β ( Fig . 1E–H ) , indicating that members of Orthopoxvirus genera , including ECTV , sustained levels of phospho-IκBα and inhibited the degradation of IκBα . Since IκBα appeared to be phosphorylated but not rapidly degraded during infection with ECTV or VACV , we sought to determine if NF-κB p65 was retained within the cytoplasm . HeLa cells were mock-infected or infected with ECTV or VACV , and p65 nuclear accumulation was assayed using immunofluorescence ( Fig . 2A and D ) . Mock-infected cells demonstrated cytoplasmic retention of p65 in the absence of TNFα or IL-1β stimulation , as expected ( Fig . 2A panels a–c and D panels m–o ) . In contrast , mock-infected cells stimulated with TNFα or IL-1β showed nuclear accumulation of p65 ( Fig . 2A panels d–f and D panels p–r ) . Upon infection with ECTV or VACV , p65 was retained in the cytoplasm following treatment with TNFα or IL-1β , indicating that ECTV could inhibit NF-κB despite the lack of orthologs of M2 , K7 , B14 , A49 , and A52 ( Fig . 2A panels g–l and D panels s–x ) . These data were confirmed by Western blotting cytoplasmic and nuclear extracts from infected HeLa cells with an antibody specific for p65 ( Fig . 2B and E ) . As expected , p65 was absent from the nuclear extract of mock-infected cells . In contrast , mock-infected cells treated with TNFα or IL-1β showed nuclear p65 ( Fig . 2B and E ) . Little p65 accumulation in the nuclear extract was observed in cells infected with ECTV or VACV and treated with TNFα or IL-1β ( Fig . 2B and E ) . These results were also confirmed in mouse embryonic fibroblasts ( MEF ) ( Fig . 2C and F ) . Together , these data indicate that NF-κB signalling is inhibited upon infection with members of the Orthopoxvirus genus . Importantly , despite the lack M2 , K7 , B14 , A49 and A52 in ECTV , ECTV infection clearly inhibited p65 translocation to the nucleus . Recently , we identified a family of four ankyrin/F-box proteins in ECTV ( EVM002 , EVM005 , EVM154 and EVM165 ) , which interact with the cellular SCF ubiquitin ligase ( Fig . S1 ) [25] . The poxvirus family of ankyrin/F-box proteins differs substantially from the cellular F-box proteins . In contrast to the cellular F-box proteins , the poxviral F-box domains are found at the C-terminus in combination with N-terminal ankyrin repeats [25] , [26] , [29]–[34] . With the exception of EVM005 , which has only one ortholog in cowpox virus Brighton Red , CPXVBR011 , multiple orthologs exist for EVM002 , EVM154 and EVM165 . Since the SCFβ-TRCP ubiquitin ligase plays an essential role degrading phospho-IκBα , we sought to determine if EVM005 could inhibit IκBα degradation and NF-κB activation during ECTV infection [27] , [28] . We first tested the ability of EVM005 to inhibit the nuclear accumulation of NF-κB p65 . HeLa cells were mock-transfected or transfected with full length Flag-EVM005 . At 12 hours post-transfection , cells were stimulated with TNFα or IL-1β for 20 minutes and nuclear accumulation of p65 was detected using immunofluorescence ( Fig . 3 ) . As expected , unstimulated HeLa cells demonstrated cytoplasmic staining of p65 ( Fig . 3A panels a–c and B panels m–o ) , and strong nuclear accumulation of p65 was seen following TNFα and IL-1β stimulation ( Fig . 3A panels d–f and B panels p–r ) . In contrast , cells expressing Flag-EVM005 that were stimulated with TNFα or IL-1β strongly inhibited p65 nuclear accumulation ( Fig . 3A panels g–i and B panels s–u ) . Given the importance of the F-box domain in associating with the SCF ubiquitin ligase , we wanted to determine whether this domain also contributed to the inhibition of NF-κB activation . To do this , we utilized an EVM005 mutant , Flag-EVM005 ( 1-593 ) , which lacks the C-terminal F-box-like domain and fails to interact with the SCF ubiquitin ligase [25] . Interestingly , cells expressing Flag-EVM005 ( 1-593 ) displayed strong nuclear staining of p65 following TNFα or IL-1β stimulation ( Fig . 3A panels j–l and B panels v–x ) . Nuclear translocation of p65 was quantified by counting cells from three independent experiments ( Fig . 3C ) . These data indicate that expression of Flag-EVM005 inhibited both TNFα- and IL-1β-induced nuclear accumulation of p65 , and that inhibition of p65 nuclear accumulation required a functional F-box domain . Since transient expression of EVM005 inhibited p65 translocation ( Fig . 3 ) , we sought to determine if EVM005 stabilized IκBα . HeLa cells were transfected with Flag-EVM005 and IκBα was visualized using immunofluorescence ( Fig . 4A ) . As expected , in unstimulated cells IκBα was present within the cytoplasm ( Fig . 4A panel a–d ) . Following 20 minutes of treatment with TNFα , the level of IκBα within the cytoplasm dramatically decreased as a result of ubiquitination and degradation of IκBα ( Fig . 4A panel e–h ) [27] , [28] . Expression of Flag-EVM005 in the absence of TNFα stimulation showed that the levels of IκBα were unaffected ( Fig . 4A panel i–l ) . In contrast , cells expressing Flag-EVM005 and stimulated with TNFα demonstrated that ectopic expression of EVM005 stabilized IκBα compared to the surrounding cells ( Fig . 4A panel m–p ) . Levels of IκBα were unaffected by expression of Flag-EVM005 ( 1-593 ) ( Fig . 4A panel q–t ) ; however , upon treatment with TNFα , IκBα was degraded , suggesting that the F-box domain was necessary for EVM005 to inhibit IκBα degradation ( Fig . 4A panel u–x ) . To further confirm these data , HeLa cells were mock-transfected or transfected with Flag-EVM005 in the absence or presence of TNFα and immunoblotted for anti-IκBα , anti-Flag to detect EVM005 , and anti-β-tubulin as a loading control . Compared to unstimulated cells , which showed a significant level of IκBα , cells treated with TNFα showed decreased levels of IκBα ( Fig . 4B ) . Expression of EVM005 led to the stabilization of IκBα ( Fig . 4B ) . Finally , we tested the ability of EVM005 to inhibit IκBα degradation by flow cytometry . HeLa cells were mock-transfected , or transfected with Flag-EVM005 or Flag-EVM005 ( 1-593 ) [25] . Twenty-four hours post-transfection , cells were stimulated with TNFα or IL-1β , and fixed and stained with anti-Flag and anti-IκBα , to detect EVM005 and IκBα , respectively . Flag-positive cells were gated for analysis ( Fig . 4C panels b and e ) . Untransfected cells demonstrated levels of IκBα that were significantly decreased following TNFα or IL-1β stimulation , as indicated by a leftward shift on the histogram ( shown in green ) ( Fig . 4C panels a and d ) . Pre-treatment of HeLa cells with the proteasome inhibitor MG132 , and subsequent TNFα or IL-1β stimulation ( shown in blue ) inhibited the degradation of IκBα , as expected ( Fig . 4C panel a and d ) [27] . HeLa cells expressing Flag-EVM005 and stimulated with TNFα or IL-1β inhibited IκBα degradation ( Fig . 4C panels b and e ) ; however , expression of Flag-EVM005 ( 1-593 ) was unable to stabilize IκBα , resulting in degradation of IκBα ( Fig . 4C panels c and f ) . This experiment was repeated in triplicate and these data were quantified by measuring the percentage of cells that underwent IκBα degradation ( Fig . 4D ) . These data indicated that Flag-EVM005 strongly inhibits TNFα- and IL-1β-induced IκBα degradation , while the F-box deletion mutant failed to inhibit IκBα degradation ( Fig . 4 ) . Together , these data show that EVM005 expression blocked IκBα degradation in an F-box-dependent manner . To further examine the role of EVM005 during infection , we generated an EVM005 deletion virus , ECTV-Δ005 . In the past , deletion of open reading frames from poxvirus genomes has been performed by inserting drug selection or fluorescent markers into the gene of interest . Instead , we used the novel Selectable and Excisable Marker system that utilizes the Cre recombinase to delete the selection markers resulting a clean deletion of the targeted open reading frame [35] , [36] . A cassette containing yellow fluorescent protein fused to guanine phosphoribosyl transferase ( yfp-gpt ) was inserted into the EVM005 locus . To generate a marker-free EVM005 deletion virus , ECTV-Δ005 , we removed the yfp-gpt marker by infecting U20S cells that stably expressed a cytoplasmic mutant of the Cre recombinase ( Fig . S2 ) [36] . Additionally , two revertant viruses were generated by replacing the yfp-gpt cassette with either wild type EVM005 or EVM005 ( 1-593 ) , a mutant lacking the F-box domain . PCR amplification of the EVM005 locus from viral genomes was used to screen for the purity of our viral products ( Fig . S2 ) . Using a multi-step growth curve , no growth defects were detected upon infection with ECTV , ECTV-Δ005 , ECTV-005-rev or ECTV-005 ( 1-593 ) -rev ( Fig . S2 ) . To determine if ECTV devoid of EVM005 could still inhibit nuclear accumulation of p65 following stimulation with TNFα , HeLa cells were mock-infected , or infected with ECTV or ECTV-Δ005 . Immunofluorescence revealed that infection with ECTV-Δ005 inhibited NF-κB p65 nuclear accumulation ( Fig . 5A panels j–l ) . This was further supported by nuclear and cytoplasmic extracts in both HeLa ( Fig . 5B ) and MEF cells ( Fig . 5C ) . We also examined the effect of ECTV and ECTV-Δ005 infection on the production of NF-κB-regulated transcripts . HeLa cells were mock-infected , or infected with ECTV or ECTV-Δ005 at a MOI of 5 . At 12 hours post-infection , cells were stimulated with TNFα , and RNA samples were collected at 0 , 2 , 4 , and 6 hours post-TNFα treatment . Samples were screened for the relative levels of RNA transcripts corresponding to TNFα , IL-1β , and IL-6; genes known to be upregulated by NF-κB [37] . All samples are presented as relative units compared to GAPDH as well as the unstimulated or 0 hour time point within each sample . Mock-infected cells displayed an increase in TNFα , IL-1β , and IL-6 transcript levels at 2 hours post-TNFα stimulation , as expected ( Fig . 6 ) . Transcript levels decreased at 4 and 6 hours post-stimulation , due to the up-regulation of NF-κB inhibitors such as IκBα ( Fig . 1A ) [38] . In contrast , infection with ECTV and ECTV-Δ005 prevented transcriptional upregulation of TNFα , IL-1β , and IL-6 ( Fig . 6 ) . We additionally screened our 0 hour time points to compare basal levels of NF-κB transcripts between samples ( Fig . S5 ) . This analysis , demonstrated that basal levels of TNFα , IL-1β and IL-6 were higher in infected cells compared to mock infected cells , however , we were still unable to detect any significant changes between cells infected with ECTV versus cells infected with ECTV-Δ005 ( Fig . S5 ) . These data correlated with our previous data indicating that infection with either ECTV or ECTV-Δ005 inhibited the nuclear accumulation of NF-κB p65 . Finally , we looked upstream at IκBα levels . HeLa cells were infected with ECTV , ECTV-Δ005 , or ECTV-005-rev . At 12 hours post-infection , cells were stimulated with TNFα , fixed and stained with anti-IκBα or anti-I3L , an early poxvirus protein that is indicative of infection , and analyzed by flow cytometry . Unstimulated cells ( shown in black ) demonstrated high levels of IκBα that decreased following TNFα stimulation ( shown in green ) ( Fig . 7A panel a ) . As expected , TNFα-stimulated mock-infected cells that were pre-treated with the proteasome inhibitor MG132 maintained IκBα levels ( shown in blue ) ( Fig . 7A panel a ) . ECTV-infected cells stimulated with TNFα also indicated no change in the level of IκBα , lending further support that IκBα is not degraded in cells infected with ECTV ( Fig . 7A panel b ) . Additionally , TNFα-stimulated cells infected with ECTV-Δ005 or ECTV-005-rev also inhibited IκBα degradation ( Fig . 7A panels c and d ) . These data were quantified by measuring the percentage of cells with IκBα expression from three independent experiments to obtain standard errors ( Fig . 7B ) . Overall , despite lacking EVM005 , ECTV-Δ005 inhibits IκBα degradation . Since EVM005 is one of four ankyrin/F-box proteins in ECTV , it is possible that deletion of more than one open reading frame may be necessary to render ECTV susceptible to TNFα-induced NF-κB activation and degradation of IκBα . Therefore , we used the Selectable and Excisable Marker system to excise four open reading frames , EVM002 , EVM003 , EVM004 , and EVM005 , from the left end of the ECTV genome ( Fig . S3 and S4 ) [36] . Notably , EVM002 and EVM003 are duplicated genes that are encoded on both ends of the ECTV genome . ECTV-Δ002-005 is depleted of both copies of EVM002 , but only the left end copy of EVM003 ( Fig . S4 ) . EVM002 is an ECTV-encoded ankyrin/F-box protein that interacts with the SCF ubiquitin ligase and inhibits NF-κB activation by interacting with NF-κB1/p105 , a member of the IκB family of proteins [25] , [31] , [39] . Deletion of EVM002 from ECTV led to slightly increased NF-κB levels in vivo , contributing to decreased virulence , potentially through low level paracrine stimulation of interferon and NF-κB in neighbouring cells [40] . Significantly , deletion of the EVM002 ortholog , CPXV006 , from CPXV , rendered CPXV susceptible to NF-κB activation [39] . EVM003 encodes a vTNFR , but a copy of this gene is present at both ends of the genome . Thus , even though EVM003 was deleted from the left end of the genome , EVM003 is still expressed from the right end ( Fig . S4 ) . EVM004 encodes a BTB-only protein with unknown function [41] , [42] . We tested the ability of this virus , lacking two ankyrin/F-box proteins that inhibit NF-κB activation , to inhibit IκBα degradation . HeLa cells were mock-infected , or infected with ECTV , single deletion strains ECTV-Δ002 , ECTV-Δ005 , or the large deletion strain ECTV-Δ002-005 , and analyzed for their ability to protect against TNFα-induced IκBα degradation using flow cytometry ( Fig . 7C ) . Following stimulation with TNFα , ECTV , ECTV-Δ002 , and ECTV-Δ005 , inhibited IκBα degradation ( Fig . 7C panels j–l ) . Additionally , IκBα was still protected from degradation in cells infected with ECTV-Δ002-005 ( Fig . 7C panel m ) . As before , staining with anti-I3L indicated virus infection ( Fig . 7C panels n–r ) . These data were quantified by measuring the percentage of cells with IκBα expression from three independent experiments to obtain standard errors ( Fig . 7D ) . These data indicate that deletion of more than two ankyrin/F-box proteins , and potentially other ECTV encoded NF-κB inhibitors , may be necessary to render ECTV susceptible to TNFα-induced NF-κB activation . To determine if EVM005 was required for virulence , we used A/NCR or C57BL/6 mouse strains . A/NCR mice are highly susceptible to lethal infection by all evaluated routes , including the footpad , whereas C57BL/6 mice are only susceptible to lethal infection via the intranasal route [43]–[45] . Groups of five female C57BL/6 mice were mock-infected , or infected with 10-fold escalating doses of ECTV , ECTV-Δ005 , ECTV-005-rev , or ECTV-005 ( 1-593 ) -rev via the intranasal route with doses ranging between 102 and 104 pfu ( Fig . S6 ) . Following infection , body weight and mortality were monitored daily . Data from one challenge dose is displayed ( Fig . 8A and B ) . C57BL/6 mice infected with ECTV , ECTV-005-rev , or ECTV-005 ( 1-593 ) -rev succumbed to infection between day seven and ten; however , mice infected with ECTV-Δ005 survived through day 21 ( Fig . 8A ) . C57BL/6 mice infected with ECTV-Δ005 displayed an initial weight loss through day 13 , followed by weight gain similar to naive mice by day 21 ( Fig . 8B ) . We also assessed the contribution of EVM005 to virulence in the A/NCR mouse strain [44] , [45] . Five female A/NCR mice were mock-infected , or infected with ECTV , ECTV-Δ005 , ECTV-005-rev , or ECTV-005 ( 1-593 ) -rev via the footpad [44] . We infected sets of five mice with escalating 10-fold doses between 101 and 104 pfu per mouse and monitored daily changes in body weight , day of death and mortality ( Fig . S7 ) . Data from one challenge dose is displayed ( Fig . 8C and D ) . Similar to the data observed in C57BL/6 mice ( Fig . 8A and B ) , ECTV-Δ005 was attenuated in A/NCR mice compared to wild type ECTV , ECTV-005-rev , and ECTV-005 ( 1-593 ) -rev ( Fig . 8C and D ) . The data demonstrated that mice infected with ECTV , ECTV-005-rev , and ECTV-005 ( 1-593 ) -rev succumbed to infection by day 7 post-infection . Alternatively , two of five mice infected with ECTV-Δ005 survived through day 21 ( Fig . 8C and D ) . Together , the results suggest that EVM005 is a critical virulence factor for infection of two mouse strains by two different routes of infection . Notably , mice infected with ECTV-005 ( 1-593 ) -rev displayed similar mortality and weight loss profiles to mice infected with wild type ECTV and ECTV-005-rev , suggesting that although the F-box domain was necessary for inhibition of the NF-κB pathway in vitro , the ankyrin domains alone are sufficient for virulence . Though EVM005 was a potent inhibitor of NF-κB activation in tissue culture , deletion of EVM005 did not abrogate the ability of ECTV to prevent activation of NF-κB . Since tissue culture assays lack many components of the immune response , we wanted to explore the contribution of EVM005 to immune inhibition and virulence in vivo . To monitor virus spread and activation of the immune response , C57BL/6 mice were infected via intranasal inoculation with ECTV or ECTV-Δ005 , and sacrificed at days 3 , 4 , and 7 post-infection ( Fig . 9 ) . Tissue from the spleen , liver , lungs , and kidneys was harvested , and the amount of virus present was determined by plaque assay ( Fig . 9A and B ) . At 4 days post-infection , ECTV and ECTV-Δ005 showed no significant difference in growth rate in all organs tested ( Fig . 9A ) ; however , at 7 days post-infection , ECTV had grown to significantly higher levels than ECTV-Δ005 in lung , kidney and liver tissues ( Fig . 9B ) . Notably , the decrease in viral spread correlates well with the decreased mortality previously described ( Fig . 8A ) . To measure the immune response , whole blood and splenocytes were harvested on days 3 and 7 post-infection and immune cell populations were quantified using flow cytometry ( Fig . 9C–F ) . In mice infected with ECTV-Δ005 , there was a significant increase in circulating and splenic NK cells at day 7 compared to ECTV-infected mice ( Fig . 9C and E ) . Additionally , we observed a significant increase in circulating virus-specific CD8+ T-cells at day 7 post-infection in mice infected with ECTV-Δ005 compared to those infected with ECTV ( Fig . 9F ) . Notably , we did not observe an increase in virus-specific CD8+ T-cells in the spleen on day 7 ( Fig . 9D ) . The data suggest that the virus-specific CD8+ T-cells are being activated and expanded in non-splenic tissues before entering circulation . Finally , we assayed for transcriptional upregulation of NF-κB-regulated genes in liver and spleen on day 7 post-infection ( Fig . 9G and H ) . Transcriptional upregulation of TNFα , IL-1β , and IL-6 was determined by harvesting RNA from tissue samples and subjecting it to qRT-PCR . Mice infected with ECTV-Δ005 did not demonstrate an increase in NF-κB-regulated transcripts compared to ECTV-infected mice . These observations support a role for EVM005 in regulating virulence that is independent of its ability to inhibit NF-κB activation . Together these data indicate that mice infected with ECTV-Δ005 displayed boosted immune cell repertoires , increased viral clearance , and decreased mortality compared to mice infected with wild type ECTV .
The NF-κB family of transcription factors regulate a variety of genes involved in inflammation and innate immunity [1] . Not surprisingly , viruses have evolved multiple mechanisms to regulate NF-κB [3] , [46] , and a growing number of poxviral NF-κB inhibitors can be added to this list [5] . Previously , we identified four ankyrin/F-box proteins in ECTV that interact with the SCF ubiquitin ligase via C-terminal F-box domains; potentially recruiting a unique set of proteins to the SCF ubiquitin ligase [25] . The NF-κB signalling pathway is dependent on the SCF ubiquitin ligase for ubiquitination and degradation of the inhibitory protein , IκBα [47] . Here we demonstrate that IκBα is phosphorylated but not degraded during ECTV infection , suggesting that signalling is inhibited at the point of IκBα ubiquitination , an event mediated by the SCF ubiquitin ligase ( Fig . 1 ) . Additionally , we demonstrate that the ECTV-encoded ankyrin/F-box protein , EVM005 , inhibits p65 nuclear accumulation and IκBα degradation in a process that requires its C-terminal F-box domain ( Fig . 3 and 4 ) . From this , we conclude that EVM005 is an inhibitor of NF-κB activation through manipulation of the SCF ubiquitin ligase . ECTV lacking the EVM005 open reading frame , ECTV-Δ005 , was created and tested for its ability to inhibit NF-κB activation ( Fig . 5–7 ) . Even though EVM005 was deleted from the genome , ECTV-Δ005 still inhibited p65 nuclear accumulation ( Fig . 5 ) , the production of NF-κB-regulated transcripts ( Fig . 6 ) , and degradation of IκBα in tissue culture ( Fig . 7 ) . Significantly , ECTV lacking EVM005 was attenuated in both A/NCR and C57BL/6 mouse strains , indicating an additional NF-κB-independent mechanism for EVM0005 ( Fig . 8 ) . Interestingly , ECTV expressing a mutant of EVM005 lacking the F-box domain was still virulent , demonstrating that the ankyrin domains alone were sufficient for virulence ( Fig . 8 ) . Mice infected with ECTV devoid of EVM005 were able to mount a stronger immune response , consisting of higher numbers of NK cells and virus-specific CD8+ T-cells ( Fig . 9 ) . A strong immune response is most likely responsible for virus clearance and decreased mortality of mice , and the observed decrease in virus spread to the liver , lungs , spleen , and kidneys ( Fig . 9 ) . EVM005 is one of many open reading frames encoded by ECTV that inhibits NF-κB activation . Given the plethora of poxvirus-encoded inhibitors of NF-κB , the deletion of multiple open reading frames is likely required to render ECTV susceptible to NF-κB activation . In an attempt to create a strain of ECTV that was unable to inhibit TNFα-induced NF-κB activation , we deleted four open reading frames from the left end of the ECTV genome , including EVM002 , EVM003 , EVM004 , and EVM005 [36] . Large deletion strains of VACV , such as VACV811 , and modified vaccinia virus Ankara ( MVA ) , have been tremendous tools for the characterization of novel poxvirus-host interactions [48] , [49] . Although VACV811 is missing 55 open reading frames , this virus is capable of inhibiting NF-κB activation [50] . MVA is an attenuated strain of VACV that has been passaged over 500 times in chicken embryonic fibroblasts and has acquired numerous gene deletions , truncations , and point mutations [49] . MVA is the only large deletion virus that has been rendered susceptible to TNFα induced NF-κB activation [51] . Our large deletion strain of ECTV , ECTV-Δ002-005 , was able to inhibit TNFα-induced IκBα degradation ( Fig . 7C and D ) . EVM002 is an ankyrin/F-box protein that we have previously shown to interact with the SCF ubiquitin ligase and inhibit p65 nuclear accumulation [25] , [31] . These data suggest that deletion of more than two ankyrin/F-box proteins , and potentially other ECTV-encoded inhibitors of NF-κB activation , would be required to render ECTV susceptible to NF-κB activation . Creation of an ECTV strain unable to inhibit NF-κB activation would allow us to investigate how ECTV infection triggers NF-κB activation , since little is known about how poxviruses activate this pathway . Regulation of the NF-κB pathway by poxviruses has been investigated , and a variety of unique NF-κB inhibitors have been found [5] , [46] . These inhibitors include poxvirus-secreted proteins , such as the soluble vTNFR and vIL-1R [13]–[15] , as well as eight VACV-encoded proteins that act within the cell , including M2 , K1 , B14 , N1 , K7 , A46 , A49 , and A52 [16]–[22] , [24] . Of the known virus encoded inhibitors of NF-κB , only K1 , N1 and A46 contain orthologs in ECTV [16] , [17] , [19]–[22] . That ECTV is missing many NF-κB inhibitors is perhaps what contributes to the variation observed in phospho-IκBα accumulation between ECTV and VACV , where VACV-infected HeLa cells showed lower levels of phospho-IκBα accumulation , and accumulation was delayed in comparison to ECTV-infected cells ( Fig . 1A ) . This observation may be linked to the additional upstream inhibitors encoded by VACV . Our data demonstrate an accumulation of phospho-IκBα in ECTV-infected cells that is linked to regulation of the cellular SCF ubiquitin ligase by poxviral ankyrin/F-box proteins . The cellular F-box protein , β-TRCP , recognizes phospho-IκBα in uninfected cells and mediates ubiquitination and subsequent degradation via the 26S proteasome [52] . Though ECTV encodes four ankyrin/F-box proteins [25] , [34] , we tested the ability of EVM005 to inhibit NF-κB signalling , since it is unique to ECTV and CPXV [25] . Our data demonstrate that EVM005 inhibited IκBα degradation , perhaps through competition with β-TRCP for available Skp1 binding sites at the SCF ubiquitin ligase . This competition would disrupt the association between Skp1 and β-TRCP , an interaction that is required for IκBα ubiquitination and degradation . This idea is consistent with our data demonstrating the requirement of the F-box domain by EVM005 in order to inhibit degradation of IκBα ( Fig . 4 and 5 ) . In a similar fashion , HIV-encoded Vpu disrupts the association between β-TRCP and Skp1 , thus inhibiting the ubiquitination and degradation of IκBα [53] . Notably , the VACV protein A49 inhibits NF-κB signalling by binding to β-TRCP in a similar fashion to Vpu [24] . This represents a fascinating example of convergent evolution , since both EVM005 and A49 serve similar functions to inhibit NF-κB signalling , but by targeting different proteins within the SCF ubiquitin ligase . Notably , an EVM005 ortholog is not encoded by VACV , and A49 is not encoded by ECTV , demonstrating the importance of regulating NF-κB through the SCF ubiquitin ligase . Of note , our data do not rule out the possibility that EVM005 recruits substrates that are involved in NF-κB activation for ubiquitination; however , we were unable to detect degradation of IκBα , NF-κB1 p50/105 , or p65 in ECTV-infected cells ( N . van Buuren and M . Barry unpublished data ) . Regulation of NF-κB activation by poxviral ankyrin/F-box proteins has been investigated for ECTV-encoded EVM002 , CPXV-encoded proteins , CP77 and CPXV006 , and the variola protein , G1R [29] , [31] , [39] , [40] . Similar to EVM005 , these proteins interact with the cellular SCF ubiquitin ligase [29] , [31] . In contrast to EVM005 , the mechanism by which G1R inhibits NF-κB activation does not depend on the F-box domain [31] , [39] . Instead , G1R and its orthologs , CPXV006 and EVM002 , bind to the N-terminus of p105 , an inhibitory protein similar to IκBα , to prevent TNFα-induced degradation [31] . Degradation of p105 is generally mediated by the SCFβ-TRCP ubiquitin ligase following TNFα stimulation , similar to IκBα [54] . Deletion of CPXV006 , a G1R ortholog encoded by CPXV , rendered CPXV susceptible to NF-κB activation [39] . Additionally , ECTV lacking EVM002 demonstrated decreased virulence and slightly increased levels of NF-κB activation in vivo [40] . We demonstrated that ECTV lacking EVM002 still inhibited IκBα degradation in tissue culture , demonstrating that the ECTV ankyrin/F-box proteins act collectively to inhibit IκBα degradation ( Fig . 7C and D ) . In contrast , ECTV lacking EVM005 was still a potent inhibitor of NF-κB activation in culture and in vivo ( Fig . 5–7 , and 9 ) . CP77 contains a shortened F-box domain that is necessary to inhibit NF-κB activation [29] . Additionally , CP77 binds to free p65 through its ankyrin repeat domains . The model suggests that CP77 replaces the regulatory protein IκBα , following its degradation , holding the NF-κB transcription factor , p65 , inactive in the cytoplasm [29] . In contrast , we were unable to detect an interaction between EVM005 with p65 , p50/105 or IκBα ( N . van Buuren and M . Barry , unpublished data ) . Similar to our data with EVM005 , CP77 serves a dual role for CPXV as a host range factor in addition to its role in the inhibition of NF-κB activation [55] , [56] . It is clear that the ankyrin domains play a major role for most of the ankyrin/F-box proteins described to date , and this is consistent with the virulent phenotype of ECTV-005 ( 1-593 ) -rev ( Fig . 8 ) . Discovery of binding partners for the ankyrin domains of EVM005 will likely provide insight to the mechanism of virulence controlled by EVM005 . Together , the data suggest that the poxvirus encoded ankyrin/F-box proteins possess unique mechanisms to regulate NF-κB activation . Although poxviral ankyrin/F-box proteins associate with Skp1 in the SCF ubiquitin ligase through their F-box domains , identification of substrates recruited for ubiquitination has eluded the field . Additionally , of the sixty-nine cellular F-box proteins encoded in the human genome , substrates have been identified for only nine [57] . The poxviral F-box proteins are suspected to function as substrate adaptor molecules for the SCF ubiquitin ligase , using their unique ankyrin domains to recruit still unidentified cellular or viral target proteins . Although binding partners , other than Skp1 , have been identified for several of the poxvirus ankyrin/F-box proteins , none of these identified proteins have been characterized as bona fide substrates for ubiquitination [30] , [31] , [55] . In support of the substrate hypothesis , the ankyrin-only mutant virus , ECTV-005 ( 1-593 ) -rev , was still virulent , supporting a critical role for the ankyrin domains , potentially in substrate recruitment . The data presented in this paper suggest a mechanism in which EVM005 inhibits degradation of cellular substrates , such as IκBα . This suggests an alternative mechanism for the poxviral ankyrin/F-box proteins as inhibitors of the SCF ubiquitin ligase . Finally , we determined that EVM005 was a required virulence factor for ECTV during infection of C57BL/6 and A/NCR mice . However , ECTV-Δ005 was capable of inhibiting IκBα degradation , p65 nuclear accumulation and the synthesis of NF-κB regulated transcripts ( Fig . 5 , 6 and 7 ) . These data suggest that an EVM005 function independent of NF-κB inhibition is responsible for mediating virulence during ECTV infection . To this end , we demonstrated that ECTV-Δ005 spread in vivo was suppressed compared to ECTV and that this observation correlated with increased immune cell activation ( Fig . 9 ) . It is possible that EVM005 regulates the immune response in vivo . At this time , any regulation of the immune response appears to be independent of NF-κB activation as we were unable to detect increased transcription of TNFα , IL-1β or IL-6 in spleens or livers of infected mice on day 7 ( Fig . 9 ) . We hypothesize that EVM005 is recruiting substrates to the SCF ubiquitin ligase through its ankyrin domains . Infection of mice with ECTV-005 ( 1-593 ) -rev demonstrated that expression of the ankyrin domains alone was sufficient for virulence in both the A/NCR and C57BL/6 mice . Potentially the ankyrin-only mutant is still able to bind and sequester these hypothetical SCF substrates and that sequestration alone was sufficient for virulence . If the poxvirus-encoded ankyrin/F-box proteins truly function as substrate adaptors for the cellular SCF ubiquitin ligase , the identification of substrates through proteomics approaches could lead to insight into how EVM005 aids in virulence . Additionally , as mice infected with ECTV-Δ005 demonstrated increased immune responses , we feel that it is therefore likely that these hypothetical target substrates function in immune cell regulation . In conclusion , our data show that ECTV-encoded EVM005 is a unique inhibitor of NF-κB activation and also suggests the existence of an NF-κB-independent mechanism for EVM005 to contribute to virulence and inhibition of immune activation . In contrast to previously characterized poxviral inhibitors of NF-κB , EVM005 requires its C-terminal F-box domain to manipulate the cellular SCF ubiquitin ligase and inhibit IκBα degradation . Further characterization of the NF-κB-independent mechanism of virulence mediated by EVM005 as well as the identification of ubiquitinated substrate proteins remains a goal of our laboratory .
HeLa , mouse embryonic fibroblast ( MEF ) , and Baby Green Monkey Kidney ( BGMK ) cells were obtained from the American Type Culture Collection . U20S-Cre cells were generously provided by Dr . John Bell ( University of Ottawa , Ottawa , Canada ) . HeLa and U20S-Cre cells were cultured in Dulbecco's Modified Eagle Medium ( DMEM ) supplemented with 10% fetal bovine serum , 50 U/ml of penicillin , 50 µg/ml of streptomycin and 200 µM glutamine ( Invitrogen Corporation ) . MEF cells were cultured in Dulbecco's Modified Eagle Medium ( DMEM ) supplemented with 10% fetal bovine serum , 50 U/ml of penicillin , 50 µg/ml of streptomycin , 200 µM glutamine , and 10 µM non-essential amino acids ( Invitrogen Corporation ) . BGMK cells were cultured in DMEM supplemented with 10% newborn calf serum , 50 U/ml of penicillin , 50 µg/ml of streptomycin and 200 µM glutamine . Vaccinia virus strain Copenhagen ( VACV ) , and ectromelia virus strain Moscow ( ECTV ) were propagated in BGMK cells and harvested as previously described [58] . Construction of pcDNA3-Flag-EVM005 and pcDNA3-Flag-EVM005 ( 1-593 ) were previously described [25] . Construction of pDGloxP-EVM005KO was performed by amplification of 150 base pairs of DNA upstream and downstream of EVM005 from ECTV genomic DNA using Taq polymerase ( Invitrogen Corporation ) . The upstream region of homology was amplified with the following forward 5′- ( HindIII ) -AAGCTTCTCTACAAAGTATAATATATT-3′ and reverse 5′- ( XhoI ) -CTCGAGATATTATACATATTAGATGTG-3′ primers . The downstream region of homology was amplified using the following forward 5′- ( NotI ) -GCGGCCGCTCGT ACCCGCGAACAAAATAG-3′ and reverse 5′- ( BamHI ) -GGATCCTTTTTTATAAACGATA TTGTT-3′ primers . The 150 base pair fragments were cloned into pGEM-T ( Promega ) . The upstream region of homology was subcloned in to the pDGloxP vector using XhoI and HindIII restriction sites . The downstream region of homology was subcloned into the BamHI and NotI restriction sites , to create pDGloxP-EVM005KO . To clone pGEMT-EVM005-reverant the forward , 5′-ATCAATGGCCGTCTCGAT-3′ , and reverse 5′-AAGAAACAAGATACAAGA-3′ primers were used to amplify a 2787 bp PCR product from wild type ECTV viral genomic DNA using LongAmp Taq ( New England Biolabs ) . The resulting PCR product was subsequently cloned into pGEMT ( Promega ) . To clone pDG-loxP-EVM002KO , 150 bp of DNA at the 5′ end of the EVM002 open reading frame were amplified by PCR using Taq polymerase ( Invitrogen Corporation ) and the forward primer , 5′- ( HindIII ) -AAGCTTCTCATAATGATTTACTTTTTC-3′ and the reverse primer , 5′- ( XhoI ) -CTCGAGCGATTCCGTCCAAGATGATAA-3′ . The 150 bp of DNA at the 3′ end of the EVM002 open reading frame were amplified with the forward primer , 5′- ( NotI ) -GCGGCCGCGGTGCTATATCTTTTCCGTTT-3′ , and the reverse primer , 5′- ( BamHI ) -GGATCCTAGAAAGAAAATATTTAAAAA-3′ . The 5′ and 3′ 150 bp regions of homology were TA cloned into pGEMT ( Promega ) following PCR . The 5′ and 3′ regions of homology were then subcloned one at a time into the pDGloxPKO vector using HindIII and XhoI , followed by BamHI and NotI , for the 5′ and 3′ sides , respectively . BGMK cells were infected with ECTV at a MOI of 0 . 01 and transfected with 10 µg of linearized pDGloxP-EVM005KO or pDGloxP-EVM002KO using Lipofectamine 2000 ( Invitrogen Corporation ) [35] , [36] . Recombinant ECTV-Δ005-YFP-GPT or ECTV-Δ002-YFP-GPT were selected in BGMK media containing 250 µg/ml xanthine ( Sigma-Aldrich ) , 15 µg/ml hypoxanthine ( Sigma-Aldrich ) , and 25 µg/ml mycophenolic acid ( MPA ) ( Sigma-Aldrich ) . Drug resistance and YFP fluorescence were used to select recombinant viruses . Removal of the yfp-gpt marker cassette from ECTV-Δ005-YFP-GPT or ECTV-Δ002-YFP-GPT was performed using U20S cells stably expressing a cytoplasmic mutant of the Cre recombinase ( U20S-Cre ) ( provided by Dr . J . Bell , University of Ottawa ) . White ECTV foci , lacking YFP-GPT expression were selected and purified to create ECTV-Δ005 and ECTV-Δ002 . ECTV-005-rev and ECTV-005 ( 1-593 ) -rev were cloned by infecting BGMK cells with ECTV-Δ005-YFP-GPT at an MOI of 0 . 01 followed by transfection with pGEMT-EVM005-rev or pGEMT-EVM005 ( 1-593 ) -rev plasmids . Infected cells were harvested at 48 hours post-infection and recombinant ECTV-005-rev or ECTV-005 ( 1-593 ) -rev were selected through lack of YFP fluorescence using a fluorescent inverted microscope and a FITC filter ( Leica ) . PCR analysis of viral genomes verified insertion and deletion of the yfp-gpt cassette . Taq polymerase and forward 5′- ( HindIII ) -AAGCTTCTCTACAAAGTATAATATATT-3′ and reverse 5′- ( BamHI ) -GGATCCTTTTTTATAAACGATATTGTT-3′ primers were used to amplify the EVM005 locus . A multi-step growth curve was used to analyze the growth of ECTV-Δ005 , ECTV-005-rev and ECTV-005 ( 1-593 ) -rev on BGMK cells . BGMK cells were infected with ECTV , ECTV-Δ005 , ECTV-005-rev or ECTV-005 ( 1-593 ) -rev at an MOI of 0 . 05 to perform the multi-step growth curve . Virus growth was assayed using plaque assays from samples collected at indicated time points . To create the large deletion strain of ECTV , lacking EVM002 , EVM003 , EVM004 and EVM005 from the left end of the genome , we used sequential insertion and Cre-mediated excision of the yfp-gpt cassette ( Fig . S3 ) . Following Cre-mediated excision , one residual loxP site remains in place of the yfp-gpt cassette . BGMK cells were infected with ECTV-Δ002 at a MOI of 0 . 01 and transfected with 10 µg of linearized pDGloxP-EVM005KO using Lipofectamine 2000 ( Invitrogen Corporation ) . YFP-GPT positive virus was selected as described above to create ECTV-Δ002/Δ005-YFP-GPT . To delete EVM002 , EVM003 , EVM004 and EVM005 , U20S-Cre cells were infected with ECTV-Δ002/Δ005-YFP-GPT at a MOI of 0 . 01 and white foci were selected using immunofluorescence . The resulting virus , ECTV-Δ002-005 , lacks all sequences between the residual loxP site in the EVM002 locus and the new loxP site at the right side of the EVM005 locus introduced during recombination with pDGloxP-EVM005KO . PCR analysis of the EVM002 , EVM003 , EVM004 and EVM005 loci confirmed the identity and purity of this large deletion strain of ECTV ( Fig . S4 ) . Mouse and rabbit anti-Flag M2 were purchased from Sigma-Aldrich , anti-poly ( ADP-ribose ) polymerase ( PARP ) was purchased by ( BD Biosciences ) and anti-β-tubulin was purchased from ECM Biosciences . Antibodies specific for Skp1 and I5L were previously described [25] , [59] Antibodies recognizing the early poxvirus protein I3L were generously donated by Dr . David Evans ( University of Alberta ) . Antibodies recognizing IκBα and phospho-IκBα were purchased from Cell Signalling Technologies . Anti-NF-κB p65 was purchased from Santa Cruz Biotechnology . Antibodies that detected cell surface markers CD45 , NK1 . 1 , CD3 , and CD8 were purchased from BD Biosciences . An APC labeled tetramer specific to the immunodominant epitope for VACV B8R/CD8 T cell expression was obtained from the NIH tetramer facility . HeLa cells or MEF cells were mock-infected or infected with ECTV , VACV , or ECTV-Δ005 at a MOI of 5 for 12 hours followed by stimulation with either 10 ng/ml TNFα ( Roche ) or 10 ng/ml IL-1β ( PeproTech Inc ) for 20 minutes . Cells were harvested and lysed in cytoplasmic extract buffer containing 10 mM HEPES , 10 mM KCl , 0 . 1 mM EDTA ( pH 8 . 0 ) , 0 . 1 mM EGTA ( pH 8 . 0 ) , 1 mM dithiolthreitol ( DTT ) and 0 . 05% NP40 . Samples were centrifuged at 1000× g for five minutes to remove nuclei . Supernatants were collected and resuspended in SDS sample buffer . The nuclear pellets were washed and resuspended in nuclear extract buffer containing 20 mM HEPES , 25% glycerol , 0 . 4M NaCl , 1 mM EDTA ( pH 8 . 0 ) , 1 mM EGTA ( pH 8 . 0 ) , and 1 mM DTT and lysis was performed on ice for 30 minutes . Samples were centrifuged at 1000× g for five minutes . Supernatants were collected as nuclear extracts and mixed with SDS sample buffer . HeLa cells were mock-transfected or transfected with pcDNA3-Flag-EVM005 or pcDNA3-Flag-EVM005 ( 1-593 ) . Alternatively , HeLa cells were mock-infected or infected with ECTV , VACV , or ECTV-Δ005 at a MOI of 5 . At 12 hours post-infection or transfection , cells were stimulated with 10 ng/ml TNFα ( Roche ) or 10 ng/ml IL-1β ( PeproTech Inc ) for 20 minutes and fixed with 2% paraformaldehyde ( Sigma-Aldrich ) for 10 minutes at room temperature . Cells were permeablized with 1% NP40 and blocked with 30% goat serum ( Invitrogen Corporation ) . Cells were stained with anti-NF-κB p65 ( 1∶200 ) alone or co-stained with either anti-NF-κB p65 ( 1∶200 ) and mouse anti-Flag M2 ( 1∶200 ) , or anti-IκBα ( 1∶125 ) and rabbit anti-Flag M2 ( 1∶200 ) . Cells were stained with secondary antibodies anti-mouse-AlexaFluor488 and anti-rabbit-AlexaFluor546 at a dilution of 1∶400 ( Jackson Laboratories ) . Coverslips were mounted using 4 mg/ml N-propyl-gallate ( Sigma Aldrich ) in 50% glycerol containing 250 µg/ml 4′ , 6-diamino-2-phenylindole ( DAPI ) ( Invitrogen Corporation ) to visualize nuclei . Cells were visualized using the 40× oil immersion objective of a Ziess Axiovert 200M fluorescent microscope outfitted with an ApoTome 10 optical sectioning device ( Ziess ) . To quantify the number of cells displaying a nuclear localization of p65 greater than 50 cells were counted in three independent experiments . HeLa cells were transfected with pcDNA3-Flag-EVM005 or pcDNA3-Flag-EVM005 ( 1-593 ) using Lipofectamine 2000 ( Invitrogen Corporation ) . Alternatively , HeLa cells were mock-infected or infected with ECTV , ECTV-Δ002 , ECTV-Δ005 , ECTV-005-rev , or ECTV-Δ002-005 at a MOI of 5 . At 24 hours post-transfection or 12 hours post infection , mock-infected cells were stimulated with 10 µM MG132 . Samples were then left unstimulated or stimulated with 10 ng/ml TNFα ( Roche ) or 10 ng/ml IL-1β ( PeproTech Inc ) for 20 minutes . Cells were fixed in 0 . 5% paraformaldehyde ( Sigma-Aldrich ) for 15 minutes at 37°C . Cells were permeablized with ice cold 90% methanol for 30 minutes . Transfected cells were co-stained with rabbit anti-Flag M2 ( 1∶200 ) and anti-IκBα ( 1∶400 ) . Cells were stained with anti-rabbit phycoerythrin ( 1∶1000 ) and anti-mouse-AlexaFluor488 ( 1∶400 ) ( Jackson Laboratories ) secondary antibodies , and resuspended in PBS . Infected cells were stained with anti-I3L ( 1∶100 ) or anti-IκBα ( 1∶400 ) , followed by anti-mouse-AlexaFluor 488 ( 1∶400 ) . Data were collected on a Becton Dickinson FACScan flow cytometer and analyzed with CellQuest software . Mean fluorescence intensities were calculated for three independent experiments . Whole blood or splenocytes were harvested on days 3 or 7 post infection from C57BL/6 mice infected with ECTV or ECTV-Δ005 . Whole blood was lysed with water at a 40∶1 water to blood volume ratio for ten seconds then brought to 1X with 10X PBS . The remaining white blood cells were resuspended in PBS with 2% FBS prior to staining . Spleen tissues were disrupted with the Bullet Blender ( STL Scientific ) for ∼2 minutes at room temperature using the lowest setting in PBS . The cell suspension was pelleted and the red blood cells were lysed with BD Pharm Lyse . The remaining white blood cells were resuspended in PBS with 2% FBS prior to staining . Cells were stained for flow cytometry using Fc block and the described antibody cocktails in PBS with 2% FBS for 20–30 minutes on ice . Cells were washed twice with PBS containing 2% FBS then fixed on ice with PBS containing 2% FBS and 1% methanol free formaldehyde . Stained cells were analyzed on a BD LSRII or BD Canto . NK cells were identified as being CD45 positive , CD3 negative and NK1 . 1 positive . These are defined in the literature as NK lytic cells and can only be identified in C57BL/6 mice [60] . An APC labeled tetramer specific to the immunodominant epitope for VACV B8R/CD8 T cell expression was obtained from the NIH tetramer facility [61] . Virus-specific CD8+ T-cells were identified as CD45 positive , CD8 positive , and tetramer positive . HeLa cells were mock-infected or infected with ECTV or ECTV-Δ005 . At 12 hours post infection cells were stimulated with 10 ng/ml TNFα and RNA was harvested using Trizol according to the manufacturer's protocol ( Invitrogen Corp . ) . RNA samples were converted to cDNA using Superscript II reverse transcriptase ( Invitrogen Corp . ) . Transcript levels were analyzed by real time PCR using the following primers , TNFα forward , 5′-GGCGTGGAGCTGAGAGATAAC-3′ and reverse , 5′-GGTGTGGGTGAGGAGCACAT-3′ , IL-1β forward , 5′-TTCCCAGCCCTTTTGTTGA-3′ and reverse 5′-TTAGAACCAAATGTGGCCGTG-3′ , IL-6 forward 5′-GGCACTGGCAGAAAACAACC-3′ and reverse 5′-GCAAGTCTCCTCATCGAATCC-3′ and GAPDH forward 5′-AGCCTTCTCCATGGTGGTGAAGAC-3′ and reverse 5′-CGGAGTCA ACGGATTTGGTCG-3′ . Real time PCR was performed using the Sybr-Green master mix ( Promega ) and a MyIQ ( BioRad ) thermocycler . Data analysis was performed with IQ-5 software ( BioRad ) . Data was presented as the average of three independent experiments . Additionally , we measured transcriptional activation of TNFα , IL-1β and IL-6 in infected liver and spleen through isolation of RNA using Trizol ( Invitrogen Corporation ) as per manufacturer's protocol . RNA samples were subjected to qRT-PCR as described above to quantify transcriptional upregulation with reference to GAPDH . RNA Transcripts for EVM004 , EVM005 , EVM058 , and GAPDH are analysed as described previously [41] . cDNA was used as a template and gene-specific primers were used to amplify the last 250 nucleotides ( at the 3′ end ) of each open reading frame . Transcripts were generated with the following primers: ECTV004 forward 5′-GTTTAATATCATGAACTGCGACTATCT-3′ , and reverse , 5′-TTAATAATACCTAGAAAATATTCCACGAGC-3′ , ECTV005 forward , 5′-TAGTGGTATTAGAGAGAAATGCAATCT-3′ , and reverse , 5′-TCATTCATGTGTCTGTGTTTG-3′ , I5L forward , 5′ATGGCGGATGCTATAACCGTT-3′ , and reverse , 5′-TTAACTTTTCATTAATAGGGA-3′ . To determine the role of the ECTV encoded EVM005 in virulence we infected female C57BL/6 mice . Four to six week old female C57BL/6 mice were obtained from the National Cancer Institute ( Frederick , MD ) . Groups of five mice were infected via the intranasal route with 10-fold escalating doses of wild type ECTV , ECTV-Δ005 , ECTV-005-rev or ECTV-005 ( 1-593 ) -rev . Mice were anesthetized with 0 . 1 ml/10 g body weight with ketamine HCL ( 9 mg/ml ) and xylazine ( 1 mg/ml ) by intraperitoneal injection . Anesthetized mice were laid on their dorsal side with their bodies angled so that the anterior end was raised 45° from the surface; a plastic mouse holder was used to ensure conformity . Strains of ECTV were diluted in PBS without Ca2+ and Mg2+ to the required concentration and slowly loaded into each naris ( 5 µl/naris ) . Mice were subsequently left in situ for 2 to 3 minutes before being returned to their cages . Mice were monitored for body weight daily for up to 21 days . Mice that demonstrated a drop in body weight to 70% of their original mass , or signs of severe morbidity were euthanized . To determine organ titers and immune activation , mice were sacrificed at 2 , 3 , 4 , and 7 days post-infection , and tissue from the spleen , liver , lungs , and kidney were harvested in addition to blood collected by a needle stick in the heart . Tissue was homogenized using a tissue homogenizer ( Next Advance ) , followed by dilution in PBS ( 10% w/v ) . Viral titers were determined on BSC-1 cells using a plaque assay . To prevent avoidable suffering , mice demonstrating a drop in body weight to 70% of their original mass , or signs of severe morbidity , were euthanized . Alternatively , we infected the susceptible A/NCR strain of mice to determine the role of EVM005 during an ECTV infection . Five to ten week old female A/NCR mice were obtained from the National Cancer Institute ( Frederick , MD ) . Groups of five mice with similar body mass were arranged into separate cages . Mice were anesthetized with CO2/O2 prior to infection . ECTV , ECTV-Δ005 , ECTV-005-rev , and ECTV-005 ( 1-593 ) -rev were diluted in PBS without Ca2+ and Mg2+ to the required concentration and 10 µl was used to infect mice via footpad injection . Body weight , day of death and mortality were monitored daily . Mice that demonstrated a drop in body weight to 70% of their original mass , or signs of severe morbidity were euthanized . To prevent avoidable suffering , mice demonstrating a drop in body weight to 70% of their original mass , or signs of severe morbidity , were euthanized . Mice were anesthetized with 0 . 1 ml/10 g body weight with ketamine HCL ( 9 mg/ml ) and xylazine ( 1 mg/ml ) by intraperitoneal injection . Alternatively , mice were anesthetized with CO2/O2 . Mice were euthanized by first anesthetizing them with CO2/O2 , followed by cervical dislocation . 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 . Animal experiments were performed at Saint Louis University and approved by the Institutional Animal Care and Use Committee ( #IACUC 2082 ) . Additionally , these experiments were performed in accordance with mouse ethics outlined by the Canadian Council on Animal Care and the University of Alberta .
|
Poxviruses are large dsDNA viruses that are renowned for regulating cellular pathways and manipulating the host immune response , including the NF-κB pathway . NF-κB inhibition by poxviruses is a growing area of interest and this family of viruses has developed multiple mechanisms to manipulate the pathway . Here , we focus on regulation of the NF-κB pathway by ectromelia virus , the causative agent of mousepox . We demonstrate that ectromelia virus is a potent inhibitor of the NF-κB pathway . Previously , we identified a family of four ectromelia virus genes that contain N-terminal ankyrin repeats and a C-terminal F-box domain that interacts with the cellular SCF ubiquitin ligase . Significantly , expression of the ankyrin/F-box protein , EVM005 , inhibited NF-κB , and the F-box domain was critical for NF-κB inhibition and interaction with the SCF complex . Ectromelia virus devoid of EVM005 still inhibited NF-κB , indicating that multiple gene products contribute to NF-κB inhibition . Importantly , mice infected with ectromelia virus lacking EVM005 had a robust immune response , leading to viral clearance during infection . The data present two mechanisms , one in which EVM005 inhibits NF-κB activation through manipulation of the host SCF ubiquitin ligase complex , and an additional , NF-κB-independent mechanism that drives virulence .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"animal",
"models",
"of",
"infection",
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"cell",
"biology",
"viral",
"immune",
"evasion",
"pathology",
"and",
"laboratory",
"medicine",
"virulence",
"factors",
"virology",
"pathogens",
"biology",
"and",
"life",
"sciences",
"immunology",
"microbiology",
"viral",
"diseases",
"pathogenesis"
] |
2014
|
EVM005: An Ectromelia-Encoded Protein with Dual Roles in NF-κB Inhibition and Virulence
|
The capacity of the liver to convert the metabolic input received from the incoming portal and arterial blood into the metabolic output of the outgoing venous blood has three major determinants: The intra-hepatic blood flow , the transport of metabolites between blood vessels ( sinusoids ) and hepatocytes and the metabolic capacity of hepatocytes . These determinants are not constant across the organ: Even in the normal organ , but much more pronounced in the fibrotic and cirrhotic liver , regional variability of the capillary blood pressure , tissue architecture and the expression level of metabolic enzymes ( zonation ) have been reported . Understanding how this variability may affect the regional metabolic capacity of the liver is important for the interpretation of functional liver tests and planning of pharmacological and surgical interventions . Here we present a mathematical model of the sinusoidal tissue unit ( STU ) that is composed of a single sinusoid surrounded by the space of Disse and a monolayer of hepatocytes . The total metabolic output of the liver ( arterio-venous glucose difference ) is obtained by integration across the metabolic output of a representative number of STUs . Application of the model to the hepatic glucose metabolism provided the following insights: ( i ) At portal glucose concentrations between 6–8 mM , an intra-sinusoidal glucose cycle may occur which is constituted by glucose producing periportal hepatocytes and glucose consuming pericentral hepatocytes , ( ii ) Regional variability of hepatic blood flow is higher than the corresponding regional variability of the metabolic output , ( iii ) a spatially resolved metabolic functiogram of the liver is constructed . Variations of tissue parameters are equally important as variations of enzyme activities for the control of the arterio-venous glucose difference .
In mammals , the liver is the central organ for the control of plasma glucose . To restrict variations of plasma glucose to the range between 3 mM ( after exercise or moderate fast [1] ) and 10 mM ( after meal ingestion , [2] ) , liver metabolism may switch between glucose production ( gluconeogenesis and glycogenolysis ) and glucose consumption . At the cellular level , this switch is accomplished by the interplay of multiple regulatory mechanisms including long term alterations in the abundance of metabolic enzymes , alterations in kinetic properties of key regulatory enzymes due to external hormonal regulation , alterations in enzymatic activity in response to internal substrate availability and allosteric regulation . The majority of metabolic functions of the liver are confined to hepatocytes . These cells form tightly connected cell layers that are separated from the liver capillaries ( sinusoids ) by the space of Disse [3] . The transport of plasma metabolites from the lumen of sinusoids through the space of Disse into the cytosol of hepatocytes and vice versa connects the cellular metabolism with the blood plasma . Thus , blood flow is a crucial determinant of the liver's overall metabolic output [4] . The importance to include blood perfusion and the tissue architecture into a quantitative estimate of liver metabolism has been stressed in an earlier modeling work of Chalhoub und Belovich [5] . However , this simple truth is routinely neglected in experimental and modeling studies where the metabolic capacity of the hepatocyte ( assessed in cell cultures ) is wrongly equated with the metabolic capacity of the organ . To better understand the role of blood perfusion and tissue structure for the metabolic performance of the liver , we have developed a multi-scale tissue model of the so-called sinusoidal tissue unit ( STU ) . We define a single STU by an ensemble of a single sinusoid , the accompanying space of Disse and the adjacent layers of hepatocytes . The sinusoids form the microvascular bed linking the hepatic portal vein and artery with the hepatic central vein . The exchange of substances between the blood , the space of Disse and the hepatocytes leads to a progressive alteration of plasma composition along the STU , continuously changing from the portal region ( where the blood from the liver artery and portal vein mix ) to the central region connected to the liver vein . Importantly , hepatocytes at different spatial positions of the STU display a differential endowment with metabolic enzymes . Higher amounts of the glycolytic enzymes glucokinase ( GK ) , PFK2/FBP2 , phosphofructokinase 1 ( PFK1 ) , pyruvate kinase ( PK ) are found in the pericentral region , while higher enzyme levels of the gluconeogenetic enzymes glucose-6-phosphate phosphatase ( G6PP ) , fructose1 , 6bisphosphatase ( FBP1 ) , mitochondrial pyruvate carboxylase ( PCmito ) and PEPCK are found in the periportal region [6–21] . Consistent with such a heterogeneous distribution of enzyme activities , the glycogen content of periportal and percentral hepatocytes may largely differ [22] . These findings point to a zonated carbohydrate metabolism in the liver: Periportal hepatocytes are more engaged in glucose production , while pericentral hepatocytes are generally more engaged in glucose utilization [9 , 10] . In a previous work [23] , we constructed a detailed kinetic model of liver glucose metabolism and examined the relative importance of the different modes of enzyme regulation for the dynamic behavior of hepatic glucose metabolism . However , the influence of tissue architecture and blood flow on the ability of the liver to function as glucose homeostat were not considered in this model . It is obvious that the unique sinusoidal structure allows an efficient nutrient exchange , but how alterations in this architecture impacts on liver function is largely unknown . Therefore , in this paper we combined our previously published model of the hepatocyte carbohydrate metabolism with a model of sinusoidal blood flow that comprises some important structural and functional tissue parameters as the thickness of sinusoids , width and number of endothelial fenestrae , thickness of the space of Disse , volume of hepatocytes and diffusion coefficients of organic molecules in the various compartments of the STU . We accounted for heterogeneous enzyme expression by constructing distinct metabolic models for the periportal and pericentral hepatocyte . The model was used to investigate the relative impact of metabolic zonation , tissue architecture and microperfusion on the hepatic carbohydrate metabolism .
The various hepatocytes lining the sinusoids display a heterogeneous endowment with metabolic enzymes . Fig 2C shows the reported ratios of enzyme abundances in periportal and pericentral hepatocytes . For example , the abundance of the glycolytic enzyme pyruvate kinase ( PK ) was found up to 3-fold higher in hepatocytes isolated from the pericentral region compared to cells isolated from the periportal region . For the calibration of the model , it was necessary to define a 'mean hepatocyte' ( MH ) which best represents metabolic parameters determined for the whole organ . To this end we put the maximal enzyme activities ( Vmax values ) of the MH to the arithmetic mean of the Vmax values of the hepatocytes that are spatially closest to the portal and central pole ( in the following referred to as periportal and pericentral hepatocytes , PPH and PCH ) . The ratios of the Vmax values of the PPH and PCH were put to the measured ratios of protein abundances . All Vmax-values are given in S1 Supplement . For the enzyme endowment of the individual hepatocytes lying between the PPH and PCH we assumed that the protein abundances of the key regulatory enzymes ( see Fig 2C ) change linearly from the portal to the central region , i . e . the protein abundance of the enzyme in the i-th hepatocyte is given by the relation E ( i ) =EPP+i−1N−1 ( EPC−EPP ) =EPP ( 1+i−1N−1[α−1] ) where i numbers the spatial position of the hepatocyte along the sinusoid ( i = 1 labels the hepatocyte closest to the portal pole , i = N labels the hepatocyte closest to the central pole ) , EPP and EPC are the enzyme abundances of the enzyme in the first and the last cell and α=EPCEPP is the ratio of enzyme abundances in periportal and pericentral cells . We assume that the metabolism of the whole liver can be best represented by the metabolism of a 'mean' hepatocyte at position i = N/2 which is endowed with the mean protein abundance ( EPP+ EPC ) /2 . Hence , the Vmax values of the enzymes in this cell , Vmax ( N/2 ) , were chosen such that that the best concordance between simulated metabolite concentrations and fluxes and measured values in the whole organ was achieved . Once a numerical value for Vmax ( N/2 ) is known , the Vmax values of the enzyme in the other cells can be calculated from the relation Vmax ( i ) Vmax ( N2 ) =E ( i ) E ( N2 ) because the maximal enzyme activity of an enzyme is up to a constant factor ( the turnover rate constant ) proportional to the protein abundance . First we simulated the glucose exchange flux for the PPH , PCH and MH at plasma glucose concentrations in the physiological range between 3 mM and 12 mM ( Fig 2A ) and compared them to the experimentally determined whole liver glucose exchange fluxes ( see [23] ) for details ) . The feasibility of the calculated metabolic states was checked by the good concordance of calculated and experimentally determined ranges of metabolite concentrations ( Fig 2B ) . The simulations show that the PPH has a higher capacity for gluconeogenesis , while the PCH has a higher capacity for glycolysis for each plasma glucose concentration . The set-point , i . e . the plasma glucose concentration at which net glucose exchange flux of the liver equals zero , is shifted from 7 . 0 mM for the PPH to 5 . 4 mM for the PCH . Notably , we found a similar set-point shift between hepatocytes in the fasted and fed state for the liver ( [23] ) . Next we simulated the dynamic changes of intrahepatic glycogen during a starvation-refeeding cycle ( Fig 2D ) . Experimental data of the temporal glycogen content were taken from [25] where fasted rats were fed ad libitum for 20 hours and subsequently starved . We started the simulation with a pre-fasted liver ( plasma glucose concentration of 4 mM ) . Interestingly , both the PPH and the PCH display smaller variations of the glycogen pool compared to the MH . For the calibration of the blood flow model we compared calculated time courses of various indicators with the experimental data obtained by the indicator dilution technique in an in situ perfused phenobarbital-treated rat liver [34] ( Fig 3A–3C ) . A soluble indicator substance is injected into the portal vein and the temporal concentration change in the central vein ( = indicator dilution curve ) is measured . Depending on blood flow velocity , diffusional exchange rates and the accessible distribution space , different indicators give rise to different dilution curves . Indicators used for model calibration were red blood cells ( RBC ) , albumin and water . Since RBCs are larger than the size of the fenestrae , they are confined to the vascular compartment . Albumin is small enough to penetrate the fenestrae and to enter the space of Disse , but the endocytotic uptake rate of albumin into hepatocytes is negligible within the short time window ( 1–2 minutes ) . Water can enter the space of Disse and the cellular compartment . From the average length of sinusoids ( ≈ 300 μm ) and the mean blood flow velocity one can estimate that the transition time of red blood cells trough a liver sinusoid is about 1 second . However , the experimentally determined dilution curve for red blood cells has a width of about 40 seconds . This delay is due to the time required by the red cells to be transported from the injection site to the sinusoids and from the sinusoids to the collecting veins . To take this into account , we approximated the temporal plasma profile of all indicators by the mean dilution curve of the red cells . To take into account random variations in tissue structure and blood pressure , we simulated indicator dilution curves for a multitude of liver sinusoids while randomly sampling the trans-sinusoidal pressure difference Δp and the structural parameters vessel diameter and thickness of the space of Disse from the respective observed distributions ( see Fig 3D–3F ) . The largest variability of the indicator dilution curves was obtained for labeled water where the half-life decay time varied between 50–70 seconds . The smallest variability resulted for the dilution curves of red cells . During the passage of blood through the liver , the plasma concentrations of glucose and hormones are continuously changing . We simulated the intra-sinusoidal alterations in plasma glucose and hormone concentrations at different fixed portal glucose concentrations in the physiological range of 3–15 mM . The corresponding portal hormone concentrations were calculated by means of the glucose-hormone transfer ( GHT ) function ( see S3 Supplement ) . The concentration of plasma lactate , which in our model represents all gluconeogenetic precursor substrates , was kept at 2 mM . Simulations were run until a steady state was reached . The simulations were repeated 100 times with randomly and independently sampled structural , morphological and metabolic parameters . Structural and morphological parameters were drawn from their measured distributions ( Fig 3D–3F ) , while protein abundances were sampled from equal distributions reflecting the variability of the periportal to pericentral protein ratios ( Fig 2C ) . A convenient measure to assess the contribution of the liver to the homeostasis of the plasma glucose is the portovenous glucose concentration difference . The mean and variance of the computed and the clearance of the hormones insulin and glucagon are shown in Fig 4 . In agreement with the experimental data , the model simulations predict that the liver acts as glucose producer for portal plasma glucose concentrations below 8 mM , whereas it turns to a glucose consumer for portal glucose concentrations higher than 8mM . The hormone extraction curves in Fig 4B and 4C show that during one passage a fraction of 20–40% glucagon and 30–80% of insulin can be removed by the liver . These values are also in good agreement with experimental findings [20] . Interestingly , the set-point for the STU lies at approximately 7 . 5 mM and is therefore even higher than the set-points of the isolated periportal and pericentral hepatocyte ( see Fig 2A ) . While the set-points for the individual hepatocytes were calculated with the GHTF given in appendix 3 , the relation between plasma glucose and plasma hormones changes along the portal-central axis as glucose is exchanged ( either taken up or produced ) and hormones are gradually cleared . Therefore the actual relation between glucose and hormones ( and thereby the position of the set-point ) changes from cell to cell and eventually determines the set point of the STU as a whole . The grey-shaded areas in Fig 4 indicate the range of variability in the arteriovenous glucose difference for a single STU as caused by random variations of tissue parameters . In the fibrotic and cirrhotic liver , striking structural changes are an enlargement of hepatocytes ( [44] ) and a decrease in the effective length of sinusoids ( mainly due to sinusoidal capillarization ( [45] ) . We thus simulated how the metabolic input-output relation of a single STU is affected by the length of the sinusoid ( see Fig 5 ) . This analysis revealed that the length of the sinusoid has severe consequences for the functional output . The longer the sinusoidal length–and thereby the contact time for the exchange of glucose–the higher the alteration in plasma glucose concentration . At high glucose levels a longer sinusoid clears more glucose form the blood ( red lines ) , while at low glucose levels a longer sinusoid produces more glucose ( blue lines ) . If the plasma glucose is close to the set point , the sinusoidal length is not a major determinant for glucose exchange ( green lines ) . We further analyzed the distribution of glucose within the various compartments of the STU and the contribution of individual hepatocytes to the net glucose balance . Depending on the portal glucose concentration , the glucose concentration along the sinusoid decreases or increases in a nonlinear manner ( Fig 6A ) . Intriguingly , for portal glucose concentrations in the range between 6 and 8 mM , upstream hepatocytes in the region around the portal field act as glucose producer ( = positive glucose exchange flux ) whereas downstream hepatocytes closer to the central pole act a glucose consumers ( Fig 6B ) . This gives rise to an intra-sinusoidal glucose cycle where the glucose produced by the upstream hepatocytes are reutilized by the downstream hepatocytes . As shown in Fig 6C , the glucose concentration gradient between the space of Disse and the sinusoidal lumen is largest at high portal glucose concentrations . But even in this situation the gradient remains below 1 mM . Thus , diffusion through endothelial fenestrae is sufficiently fast to prevent larger concentration gradients . Somewhat larger concentrations differences of up to 2 mM may occur between the space of Disse and the cytosol of hepatocytes ( Fig 6D ) . Next , we simulated the impact of a starvation-refeeding cycle on the dynamic changes of the glycogen content in the individual hepatocytes of the STU ( Fig 7 ) . As for the single cell case ( Fig 2D ) , the simulation started with a pre-fasted liver ( portal plasma glucose concentration of 4 mM ) where the glycogen stores of all cell are almost completely emptied . During the feeding phase , the portal glucose concentration is increased to 8 mM , before it is reset to 4 mM during the fasting period . Simulations were repeated 100 times , again randomly and independently varying structural and morphological parameters as well as metabolic enzyme abundances from their respective distributions ( see Fig 3 , Fig 2 and S1 Supplement ) . The amplitude of variations of the cellular glycogen content decrease from PPH to the PCH . In the PPH , the synthesis and degradation of glycogen PPH occurs with a significantly higher rate compared to the PCH . This implicates for hepatocytes of the periportal zone a higher glycogen content under euglycemic conditions but lower glycogen content in hypoglycaemia . To check the relative influence of metabolic parameters and tissue parameters on the performance of the STU , we carried out a sensitivity analysis of the model ( Fig 8 ) . The sensitivity S ( p ) was quantified by the change of the hepatic arteriovenous glucose difference ΔGluPV elicited by a small ( = 5% ) change of the model parameter p ( enzyme activity , tissue parameter , capillary blood pressure ) , i . e . S ( p ) =pδpδΔvPVΔvPV=1ε[ΔvPV ( p+εp ) −ΔvPV ( p−εp ) ] with ε = 0 . 05 . The relative impact of individual metabolic enzymes is dependent on the metabolic state ( fasted , fed ) , similar as found in [23] for the single hepatocyte . In contrast , the influence of structural parameters is less dependent on the metabolic state . The size of hepatocytes and the diameter of the sinusoids turned out to have the largest impact on ΔGluPV . Intriguingly , the sensitivity analysis shows that variations of the tissue architecture have an equally important influence of the metabolic performance of the liver than variations of protein abundances . The blood flow within different regions of the liver may vary up to a factor of three ( see e . g . [46 , 47] ) . We used our model to study the metabolic consequences of such variations of the regional blood flow . Blood flow was measures by perfusion CT ( see Methods ) . Fig 9 shows the CT—perfusion data and the corresponding model-based regional glucose production or utilization rate of a normal human liver . The blood flow through the sinusoid of the STU representing a small volume of liver was obtained by dividing the blood flow assessed by the perfusion CT through number of sinusoids 1 . 5 ∙ 109 ( sinusoids per 100 ml liver volume , see legend to Fig 9 ) . The average liver perfusion rate in this example was 44 . 3 ml/100ml/min with a standard variance of 6 . 1 ml/100ml/min . The computed regional glucose production rates ( hypoglycemic case: portal glucose concentration set to 4 mM ) and glucose uptake rates ( hyperglycemic case: portal glucose concentration set to 10 mM ) add up to a whole liver uptake rate of—131 . 8 μmol/g/h and production rate of 75 . 8 μmol/g/h , respectively . Interestingly , the variance of the glucose exchange fluxes is smaller than the variance of the blood flow values ( compare the histograms in Fig 9A with those in Fig 9B and 9C ) . This is a consequence of the reciprocal relationship between blood flow rate and the exchangeable fraction of metabolites between blood and tissue . A decrease of the regional blood flow reduces the venous output volume but on the other hand increases in the leaving blood the concentrations of metabolites secreted by hepatocytes and increases the concentration of metabolites cleared by hepatocytes owing to the increased time span available for the exchange of blood metabolites with the tissue . Hence the mass output ( = flow volume x concentration ) is moderately buffered against changes of the blood perfusion rate .
In summary , we presented a model of sinusoidal glucose metabolism by extending a previously established model of hepatic glucose metabolism to include liver zonation . We developed a realistic sinusoidal blood flow model taking morphological and systemic parameters into account thereby bridging the scale from the cellular to the tissue level . Taking into account variations in the tissue parameters as well as variations in the enzyme abundance we assessed the variability of sinusoidal glucose input-output relationship occurring among different livers sinusoids thereby assessing inter-organ variability . Swelling , cirrhosis and restriction of blood flow contribute to progression of liver disease thus allowing for interesting future uses of the model as a valuable tool to analyze liver functionality especially under disease conditions .
|
Glucose homeostasis is one of the central liver functions . The liver extracts glucose from the blood when plasma glucose levels are high and produces glucose when plasma glucose levels are low . To fulfill this function the liver is organized in smallest functional units , the sinusoidal tissue units ( STUs ) . These STUs consist of a single sinusoid surrounded by linear arranged hepatocytes . Liver zonation describes the spatial separation of metabolic pathways along the STUs . As blood flows through the sinusoid the plasma nutrient and hormone composition changes and in conjunction with the heterogeneous endowment of metabolic enzymes this leads to big differences in the metabolic performance of hepatocytes depending on their position within the sinusoid . This makes liver zonation and blood flow two central determinants for the functional output of the liver . In this work we present a tissue model of hepatic carbohydrate metabolism that combines liver zonation and microperfusion within the STU . We show that structural properties , enzymatic properties and regional bloodflow are equally important for the understanding of liver functionality . With our work we provide a true multi-scale model bridging the scale from the cellular to the tissue level .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion"
] |
[
"carbohydrate",
"metabolism",
"medicine",
"and",
"health",
"sciences",
"liver",
"body",
"fluids",
"chemical",
"compounds",
"enzymology",
"carbohydrates",
"glucose",
"metabolism",
"organic",
"compounds",
"glucose",
"glycobiology",
"enzyme",
"metabolism",
"enzyme",
"chemistry",
"research",
"and",
"analysis",
"methods",
"mathematical",
"functions",
"animal",
"cells",
"mathematical",
"and",
"statistical",
"techniques",
"hepatocytes",
"chemistry",
"blood",
"plasma",
"blood",
"flow",
"biochemistry",
"blood",
"organic",
"chemistry",
"cell",
"biology",
"anatomy",
"glycogens",
"physiology",
"monosaccharides",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"metabolism",
"sine",
"waves"
] |
2018
|
A multiscale modelling approach to assess the impact of metabolic zonation and microperfusion on the hepatic carbohydrate metabolism
|
Severe fever with thrombocytopenia syndrome ( SFTS ) is an emerging epidemic infectious disease caused by the SFTS bunyavirus ( SFTSV ) with an estimated high case-fatality rate of 12 . 7% to 32 . 6% . Currently , the disease has been reported in mainland China , Japan , Korea , and the United States . At present , there is no specific antiviral therapy for SFTSV infection . Considering the higher mortality rate and rapid clinical progress of SFTS , supporting the appropriate treatment in time to SFTS patients is critical . Therefore , it is very important for clinicians to predict these SFTS cases who are more likely to have a poor prognosis or even more likely to decease . In the present study , we established a simple and feasible model for assessing the severity and predicting the prognosis of SFTS patients with high sensitivity and specificity . This model may aid the physicians to immediately initiate prompt treatment to block the rapid development of the illness and reduce the fatality of SFTS patients .
Severe fever with thrombocytopenia syndrome ( SFTS ) is an emerging infectious disease caused by a novel bunyavirus , which first emerged in China in 2005[1] . In 2009 , the novel bunyavirus , designated as SFTS virus ( SFTSV ) or Huaiyangshan virus , was first isolated from patients with SFTS in China [2] . By the end of 2015 , the disease was reported in 23 provinces in China , Japan and Korea [3–5] . Another similar phlebovirus , named Heartland virus , was reported as the cause of two cases of severe febrile illness with thrombocytopenia in Missouri , USA [6] . SFTS is characterized by fever , thrombocytopenia , leukocytopenia , elevated alanine and aspartate amino transferases , proteinuria and various other systemic manifestations including muscular symptoms , respiratory symptoms , gastrointestinal symptoms , neurological disorders and coagulopathy [7 , 8] . It has become a significant public health threat with an estimated high case-fatality rate of 12 . 7% to 32 . 6% [5 , 9 , 10] . More importantly , previous research found that ribavirin may be effective more or less , but there is no specific antiviral therapy for SFTSV infection [11–15] . Symptomatic treatment and supportive therapy including intensive monitoring are the most essential part of case management [14] . Considering the higher mortality rate and rapid clinical progress of SFTS , supporting the appropriate treatment at early stage of disease and asking the doctor of intensive care unit ( ICU ) for consultation in time to SFTS patients is critical . Therefore , it is very important for clinicians to predict these SFTS cases who are more likely to have a poor prognosis or even more likely to decease . Previous studies had reported many risk factors associated with fatal outcomes [5 , 7 , 8 , 11 , 16 , 17] . For example , senior people is more likely to have a fatal clinical outcome [18] . Besides , the risk factors in terms of the clinical presentations includes acute lung injury or acute respiratory distress syndrome , central nervous system ( CNS ) symptoms , hemorrhagic manifestations , and disseminated intravascular coagulation [8 , 17] . The risk factors regarding laboratory parameters include a higher serum viral load; the imbalance of cytokines and chemokines; decreased white blood cell counts ( WBC ) and platelets ( PLT ) , higher aspartate aminotransferase ( AST ) , alanine aminotransferase ( ALT ) , lactate dehydrogenase ( LDH ) , creatinine kinase ( CK ) , alkaline phosphatase ( ALP ) , gamma-glutamyl transpeptidase ( GGT ) , blood urea nitrogen ( BUN ) level , and serum creatinine ( sCr ) and prolonged activated partial thromboplastin time ( APTT ) [5 , 7 , 8 , 11 , 16 , 17] . However , their results were not consistent . A simple , practical and accurate prognostic scoring system will be more helpful to predict the prognosis of SFTS disease . To our knowledge , only two models had been formed up today [8 , 16] . However , their scoring criterion were not consistent . The model established by Xiong et al included a subjective parameter , which limited the clinical use of this model [8] . Further , two models do not have coagulation parameters such as APTT , which is a high risk factor for fatal outcomes reported in the previous studies [5 , 8 , 16] . So , the aim of our study was to establish a simple and feasible scoring system for assessing the severity and predicting the prognosis of SFTS patients with objective parameters , through fully understanding the clinical features and disease progress of SFTS .
A total of 142 SFTS patients who were admitted to Nanjing Drum Tower Hospital , Jiangsu , between October 2010 and July 2017 were enrolled in our study . In this study , all SFTS cases were diagnosed according to the following 2 conditions: ( 1 ) acute fever of > 38°C with thrombocytopenia ( platelet ( PLT ) count <100×109/L ) , ( 2 ) laboratory-confirmed SFTSV infection by using a certificated real-time polymerase chain reaction ( RT-PCR ) kit , performed by Jiangsu provincial center for disease control and prevention [15] . The demographic factors , date of illness onset , date of admission , date of death , disease outcome , clinical presentations , physical examination and laboratory parameters of these patients were retrospectively collected . The survival cases were followed for 30 days from onset of disease . Categorical variables were summarized as frequencies and proportions . Continuous variables with a normal distribution were described as means and standard deviations ( SD ) , while the continuous variables with an abnormal distribution were shown as median and interquartile range ( IQR ) . The chi-square test was used to compare categorical variables . Two-sample t-tests were used to compare the continuous variables with a normal distribution and Mann-Whitney U-tests were used to compare the continuous variables with an abnormal distribution between fatal and non-fatal cases . The risk factors for mortality in patients with SFTS were analyzed by binary logistic regression . Variables having P values <0 . 01 in the univariate analysis were used for a multivariate stepwise logistic regression analysis . Binary logistic regression analyses were used to develop the predictive models of death of SFTS . The predictive value of the model was evaluated by the receiver operating characteristic ( ROC ) curve . Differences between the AUCs were tested using the z-test . The probability cut-off points for the optimal combination of sensitivity and specificity were determined by the Youden index . The predictive model was validated by the standard diagnostic analysis of sensitivity , specificity and positive and negative likelihood ratio ( LR ) . The patients were divided into four groups according to the interquartile range of the model ( M ) value of each patient . Kaplan-Meier survival analysis was used to compare the cumulative risk for death in the four groups , and the significance of difference was tested with the log-rank test . A P value < 0 . 05 was considered to be statistically significant . All statistical analyses were performed using SPSS ( Statistical Package for the Social Sciences ) version . 22 . 0 software ( SPSS Inc , Chicago , IL , USA ) or SigmaPlot version 12 . 5 ( Systat Software Inc . , San Jose , CA , United States ) . Patients all gave written consent to the participation in our study . The study was approved by the Ethics Committee of Nanjing Drum Tower Hospital .
A total of 142 SFTS patients were included in this study , including 33 fatal cases and 109 survival cases . The SFTS patients consisted of 67 males and 75 females . There was no difference in gender between fatal and survival cases . The median age of the fatal cases was significantly higher than that of the survival cases ( 65 . 2 vs 56 . 2 years , respectively; P<0 . 0001 ) . The median time from the onset of illness to the hospital visit was 8 days ( interquartile range [IQR] , 6–9 days ) . There was no difference about the time from the onset of illness to the hospital between fatal and survival cases . 135 patients described a history of field exposure at mountainous or hilly areas , and 7 cases had contacted the blood or body fluid of index patient within 2–3 weeks of the disease onset . The most frequently observed symptoms through the entire course and laboratory parameters on admission are shown in Tables 1 and 2 . Among these commonly presented symptoms , high fever ( T>39°C , 52% vs . 32 . 1% , P = 0 . 043 ) , respiratory symptom ( 66 . 7% vs 35 . 8% , P = 0 . 002 ) , neurologic symptoms ( 84 . 8% vs . 23 . 9% , P<0 . 0001 ) , hemorrhagic manifestations ( 48 . 5% vs . 21 . 1% , P = 0 . 002 ) and hematuria ( 93 . 1% vs . 69 . 9% , P = 0 . 011 ) were significantly overrepresented in fatal cases . Compared to patients with SFTS who survived , the PLT , natural killer ( NK ) cell and serum albumin ( Alb ) were identified to be significantly lower in deceased patients on admission , whereas the viral load , red blood cell volume distribution width ( RDW ) , B lymphocyte proportion , AST , ALT , ALP , GGT , direct bilirubin ( Dbil ) , BUN , sCr , and LDH and CK values were significantly higher , and APTT and thrombin time ( TT ) were markedly extended in deceased cases . The dynamic changes of 12 clinical laboratory parameters , including hematological and biochemical parameters , were analyzed ( Fig 1 ) . The clinical course of SFTS infection has been divided into three distinct stages: fever stage ( day 0–6 ) , multiple organ dysfunction ( MOD ) stage ( day 7–13 ) and convalescence stage ( after day 13 ) [7 , 19] . During the fever stage , thrombocytopenia , leukocytopenia , normal Alb and Hb , the elevated serum tissue enzymes level ( ALT , AST , CK , and LDH ) , the prolonged APTT and TT were observed ( Fig 1 ) . The levels of AST and LDH were significantly higher in fatal cases in comparison with no-fatal cases ( P = 0 . 034 and P = 0 . 014 , respectively ) ( Fig 1 ) . The APTT and TT was significantly longer in fatal patients when compared with survival cases ( P = 0 . 046 and P = 0 . 033 , respectively ) ( Fig 1 ) . The BUN and sCr of the survival cases were in the normal range at the first stage , while they were outliers in the fatal cases . Compared to the no-fatal cases , the BUN and sCr were significantly higher in fatal cases ( P = 0 . 005 and P = 0 . 036 , respectively ) ( Fig 1 ) . During the MOD stage , as shown in Fig 1 , an increased level of PLT in survivors but a decreased level of PLT in the patients who died was observed . The peripheral WBC counts were elevated , while Alb and Hb levels were diminished in both fatal cases and no-fatal cases during this stage . In SFTS patients who survived , the serum tissue enzymes ( AST , CK , and LDH ) began to decline slowly , but the ALT level rose slightly . However , all of them appeared to progressively rise during the MOD stage and were significantly higher in fatal cases . sCr level was increased in the fatal cases and was slowly decreased in no-fatal cases . The BUN level appeared to decline in two groups . The coagulation tests showed that the APTT and TT prolonged significantly in fatal cases , while both began to slowly shorten in survival cases ( Fig 1 ) . The Alb were markedly lower in fatal cases in comparison to survival cases ( P<0 . 0001 ) . Compared to survivors , the serum tissue enzymes ( ALT , AST , CK , and LDH ) , BUN level and sCr level were significantly higher and the APTT and TT were pronouncedly longer in deceased cases ( P = 0 . 001 , P<0 . 0001 , P<0 . 0001 , P<0 . 0001 , P<0 . 0001 , P<0 . 0001 , P<0 . 0001 and P<0 . 0001 , respectively ) ( Fig 1 ) . After day 13 , most clinical parameters converted to normal physical ranges in survivors . However , the fatal cases had the higher serum tissue enzymes and the renal function index ( BUN and sCr ) , the lower level of PLT and Alb , and prolonged APTT at this stage , as shown in Fig 1 . Univariate regression analyses of objective parameters for mortality of SFTS patients on admission were performed . The older age ( OR , 1 . 084; 95% CI , 1 . 038–1 . 133; P< 0 . 0001 ) , RDW ( OR , 1 . 852; 95% CI , 1 . 144–2 . 998; P = 0 . 012 ) , ALT ( OR , 1 . 004; 95% CI , 1 . 001–1 . 008; P = 0 . 016 ) , AST ( OR , 1 . 003; 95% CI , 1 . 001–1 . 004; P< 0 . 0001 ) , ALP ( OR , 1 . 008; 95% CI , 1 . 003–1 . 013; P = 0 . 002 ) , GGT ( OR , 1 . 003; 95% CI , 1 . 000–1 . 006; P = 0 . 020 ) , Dbil ( OR , 1 . 032; 95% CI , 1 . 004–1 . 062; P = 0 . 025 ) , Alb ( OR , 0 . 871; 95% CI , 0 . 786–0 . 965; P = 0 . 008 ) , BUN ( OR , 1 . 408; 95% CI , 1 . 236–1 . 602; P<0 . 0001 ) , sCr ( OR , 1 . 023; 95% CI , 1 . 013–1 . 033; P<0 . 0001 ) , LDH ( OR , 1 . 001; 95% CI , 1 . 001–1 . 002; P<0 . 0001 ) , APTT ( OR , 1 . 113; 95% CI , 1 . 069–1 . 159; P<0 . 0001 ) , and TT ( OR , 1 . 040; 95% CI , 1 . 024–1 . 055; P<0 . 0001 ) were the risk factors for fatal outcomes ( Table 3 ) . Variables having P values <0 . 01 in the univariate analysis were used for a multivariate stepwise logistic regression analysis . Multivariate regression analyses indicated that the older age ( OR , 1 . 117; 95% CI , 1 . 046–1 . 194; P = 0 . 001 ) , BUN level ( OR , 1 . 277; 95% CI , 1 . 084–1 . 505; P = 0 . 003 ) , and APTT ( OR , 1 . 093; 95% CI , 1 . 041–1 . 148; P<0 . 0001 ) were the independent risk factors for fatal outcomes of SFTS patients ( Table 3 ) . As showed in the Table 3 , multivariate regression analyses revealed that the older age , BUN level , APTT were the independent risk factors for fatal outcomes . Based on these independent predictors , a regression models were derived to predict fatal outcomes for SFTS patients . The model ( M ) is as follows: M=11+e− ( −14 . 521+0 . 111×Age+0 . 245×BUN+0 . 089×APTT ) ROC analysis was performed to compare the predictive value of the Model , Age , BUN and APTT . The cut-off values and area under ROC curve ( AUCs ) of these parameters for predicting death are included in Table 4 . ROC curve is depicted in Fig 2 . The model for predicting the mortality after infection with SFTSV showed an AUC of 0 . 927 ( 95% CI: 0 . 871–0 . 964 , P<0 . 0001 ) with the optimal cut-off value of 0 . 28 , which was higher than the AUCs of age ( 0 . 726 , 0 . 645–0 . 797 , P = 0 . 0001 ) , APTT ( 0 . 870 , 0 . 803–0 . 920 , P = 0 . 025 ) and BUN ( 0 . 869 , 0 . 801–0 . 920 , P = 0 . 039 ) . The model exhibited a significantly higher AUC in the prediction of death compared to the APTT ( P = 0 . 025 ) , age ( P = 0 . 0001 ) and BUN ( P = 0 . 039 ) . Kaplan-Meier survival analysis indicated that SFTS patients with the model scores >0 . 29 were much more likely to decease ( log-rank test; χ2 = 82 . 20 , P< 0 . 0001 ) during 30 days follow-up ( Fig 3 ) . M value of the model positively correlated with the fatality rate . The hospital mortality in different ranges of M value is shown in Fig 4 . When the M value was greater than 0 . 29 , the mortality was obviously higher than the overall mortality ( 74 . 3% vs 22 . 7% , respectively ) . No patient died with the M value less than 0 . 02 . These findings suggested that the patients with M >0 . 29 were much more likely to decease after SFTSV infection .
The SFTS is an emerging infectious disease which has a mortality rate of up to 30% [2] . To better understand the disease progression of SFTS and identify the possible risk factors that are related to the fatal outcome of the patients , we systemically analyzed the risk factors between the fatal and survival SFTS cases in terms of demographic data , clinical symptoms and laboratory parameters in this retrospective study . Fatal cases presented more severe clinical symptoms than non-fatal cases . High fever ( T>39°C ) , respiratory symptoms and neurologic symptoms were significantly overrepresented in fatal cases . However , not all the fatal patients have these symptoms on admission . Some cases developed these symptoms afterwards . The dynamic course of SFTS can be divided into three distinct stages of SFTS: fever , MOD , and convalescence stages according to the previous report [15] . In the present study , the serum tissue enzymes ( ALT , AST , CK , and LDH ) , BUN level and sCr level were significantly higher and the APTT and TT were distinctly longer in fatal cases as compared with the non-fatal cases with SFTS in both fever and MOD stages . In the convalescence stages , the LDH , BUN and sCr level were also significantly higher and the APTT values were distinctly longer in fatal cases . Although the laboratory parameters were significantly different between fatal and non-fatal cases of SFTS during these stages , it is critical to identify the risk factors between the fatal and survival SFTS cases as soon as possible . The accurate prediction of the prognosis promptly may aid the clinicians to take the intervention measures in advance , control the disease progression and finally improve the prognosis for SFTS patients . We compare the laboratory parameters between fatal cases and non-fatal cases of SFTS on admission . The biochemical , hematological parameters and coagulation variables on admission were shown that the PLT and Alb were significantly lower in deceased patients , whereas the viral load , B lymphocyte ratio , RDW , AST , ALT , ALP , GGT , Dbil , BUN , sCr , LDH and CK values were significantly higher , and the APTT and TT were markedly longer in deceased cases in comparison with the survivors . Thrombocytopenia , the decreased Alb an elevated AST , ALT , LDH , ALP , GGT , and sCr level were found to be the risk factors of patient mortality by univariate analysis . However , these parameters were not found to be statically significant after multivariate analysis in this study , which is not fully consistent with the previous studies [5 , 8 , 9 , 12 , 13 , 16 , 19] . We conducted univariate and multivariate analysis for objective indicators on admission which showed significant differences between the fatal and survival SFTS cases . Three identified independent risk factors are the older age , APTT and BUN , which were critical for predicting fatal outcome for SFTS patients . According to this study , the age in fatal cases was significantly higher in comparison to survival cases . Studies of demographic characteristics have also showed that age is a critical risk factor for death of SFTS patients [3 , 5 , 9 , 18] . In our analysis , the APTT is a good factor to predict mortality for the SFTS cases , which was also found in the previous studies [5 , 12 , 20] . Overall , the APTT seemed to be the predominant factors closely related to hemorrhagic tendency . Based on the published reports and our own experience , most SFTS patients experienced MOD stage [5 , 19] . Kidney function is currently recognized as an important indicator of MOD assessment , which is also important in the assessment of hemorrhagic fever with renal syndrome being caused by Hantaan virus which belongs to bunyavirus family [21] . One study also presented impaired renal and elevated BUN in an SFTSV-infected mouse model [22] . Renal function test , as a routine checkup , is simple and rapid , and is popular at all level of medical institutions now . BUN is a parameter that can indicate impairment of renal function . Our and others researches presented that BUN level as a key indicator for assessing kidney function was a high-risk factor for fatal outcome of SFTS patients [8 , 12 , 16 , 17] . Finally , we established a risk model based on the three critical risk factors of age , APTT value and BUN level on admission , which showed relatively high predictive value with high , sensitivity and specificity ( 81 . 25% and 91 . 74% , respectively ) . The AUC of this model was 0 . 927 , which was higher than the AUCs of the age , APTT and BUN . Furthermore , the Kaplan-Meyer analysis of overall survival was performed using the IQR of the M values on admission as the cut-off . The cumulative survival rate without an adverse outcome between the four groups was significantly different ( log-rank test; P<0 . 0001 ) . Survival rate was significantly lower in patients with the M value higher than 0 . 29 compared to patients with lower M values . To our best knowledge , there were two models for predicting death for SFTS patients [8 , 16] . Xiong et al proposed a scoring formula including viral load , neurologic symptoms , respiratory symptoms , and monocyte [8] . Indeed , a large amount of researches have shown that there was relatively higher serum viral load in patients who died than survivors [19 , 23–25] . But viral load might not be a feasible predictor in clinical practice . In our cohort , the time of serum being sent to laboratory to detect SFTSV RNA varied among patients ( 5–28 days after onset of illness ) . Moreover , most hospitals , especially primary healthcare facilities , do not have the capacity to detect SFTSV RNA . Samples are transported to municipal centers for disease control and prevention ( CDC ) or provincial CDC of China for testing , which may lead to a delay in the key period for determining the condition of patients and giving treatment in time for clinicians . Moreover , neurological analysis was not a good parameter when scoring SFTS patients . Indeed , we showed central nervous system manifestation was an important risk factor which can predict death in SFTS patients independently , consistent previous reports [5 , 8 , 12 , 19] . However , the description of clinical symptoms was affected by the subjectivity and experience of doctors . In addition , a study presented that neurological symptoms was not an independent risk factor for predicting death [16] . Another model established by Wang et al include baseline age , serum AST level and sCr level [16] . However , they did not include coagulation parameters in their model such as APTT which was a high-risk factor for fatal outcomes proposed in our and previous studies [5 , 8 , 12] . In addition , several findings proposed that the interval between onset and admission have an impact on the prognosis of the disease , while there was no difference between the deceased cases and survivors in our study [11 , 26] . There are several limitations in this study . Firstly , the study is a retrospective study which would lead to bias the results . In addition , it is a single-center study and the sample size is relatively small . Thirdly , this research is lack of a validation group . In conclusion , the present study provided important insights into the progression and prognosis of SFTS using age and objective laboratory indicators . The model including age , APTT value and BUN level , an inexpensive and easily calculated index , can predict fatal outcome of SFTS patients with relatively high accuracy . This model may aid the physicians to immediately initiate prompt treatment to block the rapid development of the illness and reduce the fatality of SFTS patients . However , the accuracy and sensitivity of this model needs to be further validated with large size of SFTS patients from multiple medical centers .
|
Severe fever with thrombocytopenia syndrome ( SFTS ) is an emerging infectious disease caused by a novel SFTS bunyavirus ( SFTSV ) with an estimated high case-fatality rate . However , there is no specific antiviral therapy for SFTSV infection . Symptomatic treatment and supportive therapy are the most essential part of case management . It is very important for clinicians to identify critical patients at admission . In this study , we established a simple and feasible scoring system for assessing the severity and predicting the prognosis of SFTS patients with objective parameters . This model may help the physicians to perform intervention measures in advance , control the disease progression and improve the prognosis .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"death",
"rates",
"medicine",
"and",
"health",
"sciences",
"clinical",
"laboratory",
"sciences",
"aminotransferases",
"pathology",
"and",
"laboratory",
"medicine",
"viral",
"transmission",
"and",
"infection",
"enzymes",
"enzymology",
"microbiology",
"biomarkers",
"signs",
"and",
"symptoms",
"population",
"biology",
"viral",
"load",
"thrombocytopenia",
"proteins",
"clinical",
"laboratories",
"transferases",
"hematology",
"creatinine",
"prognosis",
"population",
"metrics",
"biochemistry",
"diagnostic",
"medicine",
"fevers",
"virology",
"biology",
"and",
"life",
"sciences"
] |
2017
|
A scoring model for predicting prognosis of patients with severe fever with thrombocytopenia syndrome
|
Rift Valley fever ( RVF ) is an arthropod-borne viral zoonosis . Rift Valley fever virus ( RVFV ) is an important biological threat with the potential to spread to new susceptible areas . In addition , it is a potential biowarfare agent . We developed two potential vaccines , DNA plasmids and alphavirus replicons , expressing the Gn glycoprotein of RVFV alone or fused to three copies of complement protein , C3d . Each vaccine was administered to mice in an all DNA , all replicon , or a DNA prime/replicon boost strategy and both the humoral and cellular responses were assessed . DNA plasmids expressing Gn-C3d and alphavirus replicons expressing Gn elicited high titer neutralizing antibodies that were similar to titers elicited by the live-attenuated MP12 virus . Mice vaccinated with an inactivated form of MP12 did elicit high titer antibodies , but these antibodies were unable to neutralize RVFV infection . However , only vaccine strategies incorporating alphavirus replicons elicited cellular responses to Gn . Both vaccines strategies completely prevented weight loss and morbidity and protected against lethal RVFV challenge . Passive transfer of antisera from vaccinated mice into naïve mice showed that both DNA plasmids expressing Gn-C3d and alphavirus replicons expressing Gn elicited antibodies that protected mice as well as sera from mice immunized with MP12 . These results show that both DNA plasmids expressing Gn-C3d and alphavirus replicons expressing Gn administered alone or in a DNA prime/replicon boost strategy are effective RVFV vaccines . These vaccine strategies provide safer alternatives to using live-attenuated RVFV vaccines for human use .
Rift Valley fever ( RVF ) is an arthropod-borne viral zoonosis . The causative agent Rift Valley fever virus ( RVFV ) belongs to the genus Phlebovirus of the family Bunyaviridae and was first discovered in the Rift Valley of Kenya in 1931 [1] . RVFV infections in livestock are characterized by an acute hepatitis , abortion and high mortality rates , especially in new born or young animals . Human infection with RVFV typically leads to a mild flu-like febrile illness . However , ∼2% of infected individuals have more severe complications , such as retinal degeneration , fatal hepatitis , severe encephalitis and hemorrhagic fever [2] . The ability of RVFV to cross geographic or national boundaries , coupled with the fact that RVFV replicates in a wide range of mosquito vectors , have raised concerns that the virus might spread further into non-endemic regions of the world . Before 1977 , RVFV circulation was not detected beyond the Sub-Saharan countries . However , since 1997 , RVFV outbreaks have occurred in Egypt [3] , Mauritania in 1987 and 1998 [4] , Saudi Arabia and Yemen [5] . In 2006–2007 , RVFV outbreaks were recorded in Kenya , Somalia and Tanzania that resulted in human infections and deaths [6] . Thus , the ability of RVFV to cause explosive “virgin soil” outbreaks in previously unaffected regions demonstrates the need for prophylactic measures for this significant veterinary and public health threat . The virus genome is composed of three single-stranded negative-sense RNA segments . The large ( L ) segment ( ∼6 . 4kb ) encodes for the RNA-dependent RNA polymerase [7] . A medium ( M ) segment ( ∼3 . 8kb ) encodes for four known proteins in a single open reading frame ( ORF ) . These include the two structural glycoproteins , Gn and Gc , and the 14kDa non-structural NSm protein and the 78kDa NSm-Gn fusion peptide [7] , [8] , [9] . The small ( S ) segment is ambisense and encodes for the 1 . 6kDa viral nucleoprotein ( N ) in genomic orientation , as well as a non-structural ( NSs ) protein in the anti-genomic orientation [7] . The nonstructural genes ( NSs and NSm ) function to suppress host antiviral responses [10] , [11] . RVFV is an important zoonotic pathogen with the potential to emerge in new areas through the spread of infected insect vectors or livestock or though intentional release as a bioterror agent . [12] . Inactivated RVFV vaccine ( TSI-GSD-200 ) have been shown to elicit protective immunity in humans [13] , however multiple booster vaccinations are required to achieve protective immunity , and perhaps most importantly , for many individuals , immunity rapidly wanes in the absence of follow-up booster vaccinations [13] . A modified live virus vaccine , based upon the Smithburn strain , is available for livestock in Africa [14] , but it can cause pathology , spontaneous abortions , and teratogenic effects [15] , [16] , furthermore , animals vaccinated with live attenuated RVFV strains cannot be differentiated from naturally infected livestock , which may preclude export of these animals to non-RVFV endemic areas . One vaccine candidate under evaluation for human use is MP12 , which is a mutagen-attenuated strain of the Egyptian RVFV isolate , ZH548 [17] . This vaccine was developed for use in both humans and livestock , with encouraging results in initial animal trials , but may cause teratology in pregnant animals [18] . In addition to the adverse effects of the live-attenuated vaccines , there are considerable safety concerns including incomplete attenuation , reversion back to virulent form during the vaccine manufacturing process . Therefore , new approaches are necessary to develop safe and effective vaccines . Given limitation of existing RVFV vaccines , there is a need to explore alternative vaccine approaches . Previous studies have shown that DNA vaccines can elicit protective anti-RVFV immunity . Studies from our group and others have demonstrated that the molecular adjuvant C3d can significantly enhance antibody responses against DNA vaccine delivered antigens [19] , [20] , [21] , [22] , [23] , [24] , [25] . C3d adjuvanticity involves C3d binding to the complement receptor 2 ( CR2 ) that is located on the surface of follicular dendritic cells ( FDC ) , B cells , and T cells in many species ( For review , see Toapanta and Ross ) [26] . C3d stimulates antigen presentation by FDCs and helps to maintain immunological B cell memory . On B cells , C3d interaction with CR2 collects molecules , such as CD19 and TAPA . CD19 has a long intracellular tail that triggers a signaling cascade that results in cell activation and proliferation . Furthermore , simultaneous C3d–CR2 ligation and surface immunoglobulin ( sIg ) by antigen , activates two signaling pathways that cross-talk and synergize to activate B cells , thereby leading to enhanced antibody secretions specifically directed to the fused antigen . Therefore , we assessed whether fusion of murine C3d to the RVFV Gn glycoprotein would result in enhanced RVFV specific immune responses in the context of DNA vaccination . Alphavirus replicon vectors are single hit vectors capable of eliciting potent systemic and mucosal immune responses against a wide range of pathogens , including hemorrhagic fever viruses , such as Lassa and Ebola [27] . Recently , we and others have demonstrated that alphavirus replicons based upon either VEE or Sindbis viruses were capable of eliciting protective anti-RVFV immune responses when the vectors expressed both RVFV glycoproteins from the RVFV M segment [28] , [29] However , there has not been a direct comparison between DNA and alphavirus-based vectors , or an assessment of whether combining these vaccine strategies results in enhanced immunity or qualitative differences in the RVFV specific immune response . Furthermore , to date , RVFV vaccination studies have focused on antibody responses , and the ability of different vaccination strategies to elicit RVFV specific T cell responses has not been evaluated . Therefore , studies were conducted to directly compare DNA vaccines expressing either Gn or Gn-C3d to alphavirus vectors expressing Gn , evaluate whether combining these vaccines in a DNA prime/replicon boost strategy provided any advantage over either vaccine on its own , and to assess the nature of the antibody and T cell response elicited by each of these vaccine strategies .
pTR600 , a eukaryotic expression vector , has been described previously [23] . Briefly , the vector was constructed to contain the cytomegalovirus immediate-early promoter ( CMV-IE ) plus intron A ( IA ) for initiating transcription of eukaryotic inserts and the bovine growth hormone polyadenylation signal ( BGH poly A ) for termination of transcription . The vector contains the Col E1 origin of replication for prokaryotic replication and the kanamycin resistance gene ( Kanr ) for selection in antibiotic media . The gene sequence encoding for the RVFV , isolate ZH548 ( Genbank DQ380206 ) , Gn glycoprotein was used to PCR amplify a soluble form of Gn ( Gn ) without the transmembrane and cytoplasmic tail ( Fig . 1A ) . The Gn gene sequence was cloned into the pTR600 vaccine vector by using unique HindIII and BamHI restriction endonuclease sites . This Gn segment encoded a region from amino acids 131 to 557 ( 427 amino acids ) and terminated in the sequence VAHCP . The vectors expressing Gn fused to three tandem repeats of the mouse homologue of C3d were cloned in frame and designated Gn-C3d , similar to constructs previously described [30] . Linkers composed of two repeats of four glycines and a serine [ ( G4S ) 2] were fused at the junctures of Gn and C3d and between each C3d repeat . Potential proteolytic cleavage sites between the junctions of Gn and the junction of C3d were mutated by ligating BamHI and BglII restriction endonuclease sites to mutate an Arg codon to a Gly codon [30] . The plasmids were amplified in Escherichia coli DH5α; purified by using endotoxin-free , anion-exchange resin columns ( Qiagen , Valencia , CA , USA ) ; and stored at −20°C in distilled H2O . Plasmids were verified by appropriate restriction enzyme digestion and gel electrophoresis . Purity of DNA preparations was determined based on the optical density ( O . D . ) at wavelengths of 260 and 280 nm . A soluble form of RVFV Gn lacking the transmembrane and cytoplasmic tail ( see above ) was introduced behind the 26S subgenomic promoter of the VEE replicon plasmid pVR21 as outlined in Figure 1B . VEE replicons expressing influenza hemagglutinin were used as negative controls . VEE replicon plasmids , as well as capsid and glycoprotein plasmids were linearized with NotI , replicon and helper transcripts were generated using mMessage mMachine T7 transcription kits ( Ambion ) , and transcripts electroporated into BHK-21 cells to package replicon particles as described previously [31] . Following packaging , the replicons underwent two rounds of safety testing to ensure that no detectable replication competent virus was present [29] , [31] at which point the replicons were concentrated by ultracentrifugation through a 20% sucrose cushion and titered using polyclonal antiserum against the VEE nonstructural proteins . Expression of the truncated RVFV Gn protein from the replicon was confirmed by western blot with a Gn specific monoclonal antibody ( RV5 3G2-1A ) generously provided by Dr . George Ludwig , USAMRIID , Ft . Detrick , Frederick , MD , USA . The human embryonic kidney cell line , 293T , was transfected ( at 5×105 cells/transfection ) with 5µg of DNA by using Lipofectamine 2000 ( Invitrogen , Carlsbad , CA , USA . ) according to the manufacturer's guidelines . Supernatants were collected and stored at −20°C . Cell lysates were collected in 500µl of 1% Triton X-100 buffer and stored at −20°C . To detect specific proteins in the cell supernatant , 1 . 5% of supernatant was diluted 1∶2 in SDS sample buffer ( Bio-Rad , Hercules , CA , USA ) and loaded onto a 10% polyacrylamide–SDS gel . The resolved proteins were transferred onto a nitrocellulose membrane ( Bio-Rad , Hercules , CA , USA ) and incubated with a 1∶5 , 000 dilution of anti-RVFV mouse sera in phosphate-buffered saline ( PBS ) containing 0 . 05% Tween 20 and 5% skim milk powder . After an extensive washing , bound mouse antibodies were detected by using a 1∶5 , 000 dilution of horseradish peroxidase-conjugated goat anti-mouse antiserum and enhanced chemiluminescence ( Amersham , Buckinghamshire , United Kingdom ) . Six-to-eight week old female BALB/c mice ( Harlan Sprague-Dawley , Indianapolis , IN , USA ) were used for inoculations . Mice , housed with free access to food and water , were cared for under U . S . Department of Agriculture guidelines for laboratory animals . Mice were anesthetized with 0 . 03 to 0 . 04ml of a mixture of 5ml of ketamine HCl ( 100 mg/ml ) and 1ml of xylazine ( 20 mg/ml ) . Gene gun immunizations were performed on shaved abdominal skin by using the hand-held Bio-Rad gene delivery system as described previously [23] , [32] , [33] , [34] . For DNA immunizations , mice were immunized with three times at three week intervals with 2µg of DNA per 0 . 5mg of approximately 1-µm gold beads ( Bio-Rad , Hercules , CA , USA ) at a helium pressure setting of 400 lb/in2 . For replicon immunizations mice were given one dose at week 6 or three doses at weeks 0 , 3 , and 6 of 1×105 infectious unit ( IU ) of replicons by foot pad route . Blood samples were collected at at weeks 0 , 2 , 5 , and 8 post-vaccination . A schematic of the vaccine regimen is listed in Table 1 . Use of animals in this study was reviewed and approved by the University of Pittsburgh Institutional Animal Care and Use Committee ( IACUC ) . The attenuated strain RVFV MP12 ( MP12 ) and ZH501 was propagated and titrated using Vero cells . A pre-titrated RVFV MP12 was inactivated with 1% beta-propiolactone to a final concentration of 0 . 1% to make a whole virus inactivated preparation ( WIV MP12 ) . To ensure complete inactivation , an aliquot of inactivated virus was used to infect Vero cells and verify the lack of cytopathic effect ( data not shown ) . BALB/c mice ( n = 5 ) received a single intraperioteneal injection ( i . p . ) of MP12 ( 1×105 PFU ) 2 weeks or 8 weeks prior to infection ( Table 1 ) . Another group of mice was administered ( i . p . ) 3 doses of the WIV MP12 vaccine ( 1×105 PFU equivalent ) . Endpoint ELISA was performed on collected serum samples to assess the anti-Gn immunoglobulin G ( IgG ) response . Briefly , plates were coated with 100µl of inactivated RVFV MP12 overnight at 4°C , blocked with 5% non-fat dry milk in PBS-T ( 1h ) at 25°C , and then extensively washed with PBS-T . Serial dilutions of mouse antisera were allowed to bind ( 1h ) and the plates thoroughly washed with PBS-T . Subsequently , the primary antisera were detected by anti-mouse IgG conjugated to horseradish peroxidase ( Bio-Rad , Hercules , CA , USA ) . The reaction was detected using tetramethybenzidine ( TMB ) substrate ( Sigma , Saint Louis , MO , USA ) ( 1 h ) at 25°C . IgG isotypes were also assessed by ELISA as previously described [23] , [35] . The secondary antibodies specific for IgG1 , IgG2a , IgG2b and IgG3 ( Southern Biotechnology , Birmingham , AL , USA ) were used at varying concentrations determined by optimization . Antibody-mediated neutralization of RVFV ZH501 was measured using plaque reduction and neutralization test ( PRNT ) [36] . Briefly , 100 plaque-forming units ( PFU ) /0 . 1 ml of RVFV ZH501 was mixed with serial two fold dilutions of heat inactivated ( 60°C for 30 min ) serum samples in 96-well tissue culture plates . Virus-serum mixtures were incubated at 4°C overnight and placed into duplicate 23-mm wells ( 0 . 1ml/well ) containing confluent monolayers of Vero cells ( 2×105 ) . Cells were incubated for 1h at 37°C and 5% CO2 and overlaid with nutrient medium containing 0 . 8% agar , 5% fetal bovine serum , 200U penicillin/ml , and 200mg streptomycin/ml . The plates were incubated at 37°C and 5% CO2 . After 4 days of incubation , cells were fixed with 10% formalin and stained with 1% crystal violet for visualization of plaques . The neutralizing antibody titer of a serum was considered positive at the highest initial serum dilution that inhibited >50% of the plaques as compared to the virus control titration . The number of anti-Gn specific murine INF-γ ( mINF-γ ) secreting splenocytes was determined by enzyme-linked immunospot ( ELISpot ) assay ( R & D Systems , Minneapolis , MN , USA ) . Briefly , pre-coated anti-mIFN-γ plates were incubated ( 25°C for 1h ) with RPMI ( 200µL ) supplemented with 10% fetal calf serum and then incubated with splenocytes ( 5×105/well ) isolated from vaccinated mice . Cells were stimulated ( 48h ) with peptides ( 15mers overlapping by 11 amino acids ) representing the ectodomain of Gn glycoprotein . IL-2 was added to all wells ( 10 units/ml ) . Control wells were stimulated with PMA ( + ) ( 50 ng ) /ionomycin ( 500 ng ) or were mock stimulated ( − ) . Plates were washed with PBS-T ( 3× ) and were incubated ( 37°C for 48h; 5% CO2 ) with biotinylated anti-mIFN-γ and incubated ( 4°C for 16h ) . The plates were washed and incubated ( 25°C for 2h ) with strepavidin conjugated to alkaline phosphatase . Following extensive washing , cytokine/antibody complexes were incubated ( 25°C for 1h ) with stable BCIP/NBT chromagen . The plates were rinsed with dH2O and air-dried ( 25°C for 2h ) . Spots were counted by an ImmunoSpot ELISpot reader ( Cellular Technology Ltd . , Cleveland , OH , USA ) . At week 8 of the study , a challenge dose containing 1×103 PFU of RVFV ZH501 were administered i . p . During challenge , mice were housed in sealed negative-ventilation bio-containment units ( Allentown Inc . , Allentown , NJ , USA ) . All manipulations with infected mice and/or samples involving RVFV ZH501 were performed under strict BSL-3 enhanced conditions . The animals were examined twice daily for visual signs of morbidity or mortality , using a lab validated scoring system as previously described [37] . Mice were observed for clinical signs that ranged from lethargy , ruffled fur , and weight loss to neurological manifestations , such as hind-limb paralysis . Mice found in a moribund condition were euthanized . Sera from vaccinated mice were diluted 1∶10 in sterile PBS and 100µl of the diluted sera was injected ( i . p . ) into new , naïve BALB/c mice . One hour following transfer , the mice were challenged ( i . p . ) with virulent RVFV ZH501 ( 1×103 PFU ) . Mice were observed daily for 8 days post-transfer for signs of morbidity and mortality . Differences in ELISA titers and virus neutralization titers between various vaccine groups were analyzed by one-way ANOVA , followed by Tukey's multiple comparison test . Analysis of results from sickness score and weight loss were assessed by two-way ANOVA tests followed by Bonferroni's post tests . Statistical results are represented in the figure by * ( P<0 . 05 ) , ** ( P<0 . 01 ) , *** ( P<0 . 001 ) . Statistical analyses were done using GraphPad Prism software .
A truncated , soluble forms of Gn from the RVFV isolate ZH548 alone or fused to three copies of murine C3d ( Gn-C3d ) efficiently secreted from cells transfected with DNA plasmid and replicon vector ( Fig . 1 ) . RVFV Gn migrated at a 45kDa molecular weight and the C3d fusion with Gn increased the molecular weight to 135kDa . After 3 vaccinations , mice vaccinated with DNA plasmid expressing Gn elicited anti-Gn antibodies ( 1∶180 ) , however , the fusion of C3d to Gn enhanced the anti-Gn antibodies ( 1∶1280 ) , while mice vaccinated with replicons expressing Gn ( Rep-Gn ) had an average anti-Gn titer of 1∶2560 ( Fig . 2 ) . There were no detectable antibodies following a single DNA vaccination ( data not shown ) . In order to determine if Gn-C3d-DNA could prime and enhance antibody titers following a Rep-Gn boost , mice were vaccinated twice with Gn-C3d-DNA and then administered a single inoculation of replicon expressing Gn . These vaccinated mice had higher anti-Gn antibody titers ( 1∶4160 ) compared to mice vaccinated with a single vaccination of alphavirus-replicon ( 1∶280 ) . Mice vaccinated the Gn-DNA only , did not elicit any detectable anti-Gn antibodies ( Fig . 3A ) . These antibody responses were comparable to mice immunized with live attenuated RVFV ( MP12 ) , but 1–2 logs lower than mice vaccinated with three doses of whole-inactivated RVFV ( WIV MP12 ) . MP12 infection elicited a mixed Th1 and Th2 response , whereas mice vaccinated with three doses of WIV MP12 had a Th2-restricted immune response ( Fig . 3E and F ) . Mice vaccinated with Gn-C3d-DNA vaccines elicited predominately IgG1 , suggesting a Th2 immune response ( Fig . 3B and D ) . In contrast , the replicons expressing Gn administered to mice three times elicited not only IgG1 , but also IgG2a and IgG2b isotypes suggesting a mixed Th1/Th2 response similar to that elicited by the live attenuated MP12 vaccine ( Fig . 3C ) . Interestingly , mice primed with Gn-C3d-DNA maintained an IgG1 isotype bias following a boost with Gn expressing replicons ( Fig . 3D ) . These titers were specific to the Gn antigen , since controls ( DNA plasmid with no insert and replicons expressing the influenza virus hemagglutinin ) did not elicit anti-Gn antibodies ( data not shown ) . At week 8 of the study , sera from mice vaccinated Gn-C3d-DNA orRep-Gn neutralized ( PRNT50 ) RVFV ZH501 , while priming mice with Gn-C3d-DNA and then boosting with Rep-Gn did not significantly enhance the neutralizing titers compared to Gn-C3d-DNA or Rep-Gn alone ( Fig . 4 ) . Mice vaccinated with the live attenuated MP12 vaccine strain had the highest neutralizing titers ( average; 1∶656–1∶736 ) regardless if the mice were vaccinated at week 0 or week 6 of the study , and they were significantly higher than sera from mice vaccinated with Gn , Rep-Gn and WIV MP12 ( p<0 . 05 ) . In contrast , serum samples collected from Gn ( 1∶22 ) vaccinated or WIV MP12 ( 1∶8 ) had low virus neutralizing titers in spite of the fact that WIV MP12 elicited very high RVFV specific antibody levels as measured by ELISA ( Figure 2 ) . Mice vaccinated with DNA and replicon vaccines were challenged with MP12 virus two weeks after last immunization and splenocytes were collected 6 days post-infection . Cells collected from mice vaccinated with Gn vaccines were stimulated in vitro with 8 overlapping pools of peptide ( 15mers with overlapping by 11 ) specific for Gn . Mice vaccinated with Rep-Gn or Gn-C3d/Rep-Gn had responses to pools B and C ( Table 2 ) , representing a stretch of 111 amino acids starting at amino acid 53 in the Gn sequence . Only mice vaccinated with Gn-C3d/Rep-Gn had splenocyte responses to pool A . No responses were recorded from any mice to pools D-G . A few spots ( 10–12 spots ) were detected following stimulation of splenocytes with an irrelevant peptide or unstimulated following in vitro re-stimulation . Mice vaccinated with DNA vaccines did not elicit cellular responses ( Table 2 ) . In addition , no spots were detected above background from splenocytes collected from naïve mice immunized with MP12 at day 6 post-infection ( data not shown ) . The peptides in these pools B and C were further analyzed to determine the peptides responsible to eliciting these responses in replicon-vaccinated mice . Using a matrix format , 4 out of 10 pools ( 5 peptides/pool ) were identified ( peptide pools II , IV , VI , VII ) ( Fig . 5A ) . From this analysis , four potential peptides ( peptide # 18 , 19 , 36 , 38 ) were identified as responsible for the vaccine elicited cellular responses . Two out of four peptides share a common amino acid sequence ( SYAHHRTLL ) predicted to be MHC class I restricted ( www . immuneepitope . org ) . A unique peptide representing this region of Gn elicited similar mINF-γ cellular immune response as the four individual peptides as indicated in Figure 5B . The mice were challenged two weeks after final vaccination with a lethal dose ( 1×103 PFU ) ofRVFV ZH501 . All the mice vaccinated with an all Gn-C3d-DNA or Rep-Gn strategy or in a DNA prime/replicon boost strategy were protected from virulent virus challenge with no body weight loss or development of clinical signs ( Fig . 6 , 7 , and Table 3 ) . Sixty percent of mice that received Gn without the molecular adjuvant C3d displayed ruffled fur and lethargy with one mouse succumbing to infection ( Table 3 and Fig . 6D ) . As expected , all the mice immunized with MP12 and then challenged with RVFV ZH501 survived lethal challenge with no clinical signs of infection ( Table 3 and Fig . 6D ) . However , mice vaccinated with WIV MP12 were not protected from challenge with all mice exhibiting reduced body weight ( Fig . 6C ) , ruffled fur , lethargy ( Table 3 ) , and all mice ultimately succumbing to infection ( Fig . 6D ) . Unvaccinated naive mice had severe signs of infection and body weight loss which resulted in all mice succumbing to infection by day 4 post-challenge ( Fig . 6D ) . Mice that received appropriate DNA and replicon controls displayed clinical signs of infection ( Fig 7A and B ) and mortality was also observed in the control groups . Pooled antiserum from each vaccinated group was transferred ( i . p . ) into unimmunized mice , which were then challenged with a lethal dose of RVFV ZH501 ( Table 4 ) . Eighty percent of mice that received sera from MP12 immunized mice survived challenge . A similar outcome was observed in the Gn-C3d group where 80% of mice survived . Sera from mice primed with Gn-C3d-DNA and then boosted with Rep-Gn or immunized with Rep-Gn protected 40% ( 2/5 ) of mice , which was similar to the mice that received sera from Gn-DNA vaccinated mice . All the mice that received sera from WIV MP12 immunized mice or mice that received sera from control immunized mice ( DNA control , Rep control , Naïve ) succumbed to virulent RVFV ZH501 infection .
One of the goals of an effective RVFV vaccine is to elicit protective neutralizing antibodies . In recent years , several RVFV vaccines strategies have been employed to elicit a potent neutralizing antibody responses [36] , [38] , [39] , [40] , [41] , [42] , [43] , however , these vaccines did not always elicit high titer immune responses that protected against lethal challenge . In addition , several of these innovative strategies may not be appropriate for human use . Early RVFV vaccine studies focused on live-attenuated and inactivated virus strategies that induce long-lasting protection [36] , [44] , [45] . However , the induction of adverse reactions may likely limit the wide spread use of live-attenuated vaccines [15] , [16] , [18] . In contrast , inactivated virus vaccines often require multiple immunizations to elicit protective immune responses [46] and there is concern that the immunity elicited by these vaccines may rapidly wane without frequent booster vaccinations . In addition to potential concerns about safety or efficacy , due to the bioterrorism potential , ability to create virgin soil epidemics and zoonotic importance of RVFV , sero-surveillance is of major importance in the international trade of animals and animal-related products . Marker vaccines make it possible to differentiate infected from vaccinated animals [47] . Diagnostic tests such as RVFV recombinant N protein based ELISA and immuno-fluorescence performed on infected or transfected cells or tissues are widely used in laboratories for RVFV diagnosis [48] , [49] , [50] , [51] . Therefore , an ideal vaccine , especially for livestock applications , would lack the RVFV N protein , which would allow differentiation between vaccinated and infected individuals . To overcome the limitations discussed above , we have developed two promising vaccine candidates based on DNA plasmid and alphavirus replicon vectors that express the virus envelope glycoprotein , Gn . Each vaccine was tested alone or in a DNA prime/replicon boost strategy formulation to elicit protective immune response against virulent RVFV infection in mice . DNA vaccines have been licensed for veterinary use [52] . However , DNA vaccines have been less effective in human clinical trials for other infectious diseases [52] , [53] . In order to enhance the antibody responses elicited by DNA vaccines , our laboratory has pioneered the use of the complement protein C3d as a molecular adjuvant [20] , [23] , [25] . Since the Gn glycoprotein is known to contain protective neutralizing epitopes [38] , [40] , [42] , [54] , we focused our efforts on characterizing whether fusion of the C3d molecule to Gn resulted in enhanced RVFV specific immunity . Mice vaccinated with Gn-C3d-DNA had high titer neutralizing antibodies compared to mice vaccinated with DNA expressing Gn alone . It remains to be determined whether this effect is solely due to C3d's function as a molecular adjuvant or whether the fusion of C3d also enhances the secretion of Gn from the cell or the protein's stability in the extracellular environment . In addition to the DNA vaccine strategy , we also used a DNA prime/alphavirus replicon boost strategy to expand the repertoire of elicited immune responses . Previously , our group used a Sindbis virus replicon vectors expressing the RVFV Gn and Gc glycoproteins , as well as the non-structural NsM protein to induce protective immune responses in mice against RVFV [29] . In this study , the replicons administered alone or in a DNA prime/replicon boost strategy elicited similar anti-Gn antibody titers , however , different subclasses of IgG were elicited by each vaccine . The isotype of the polyclonal antibody in part determines the effector functions of the anti-Gn antibodies and identifies the T helper cell bias ( required for antibody class switching ) . The predominant isotype elicited by DNA and replicon immunizations was IgG1 indicating a Th2 bias . Antibodies of the IgG2a/c and IgG2b subclass fix complement proteins C1q and C3 and can opsonize and inhibit infection . IgG2a/c binds FcγRI with high avidity facilitating enhanced uptake of virus-antibody complexes by macrophages . The predominant IgG isotype elicited by DNA vaccination was IgG1 indicating a Th2 bias . However , IgG1 , IgG2a , and IgG2b were detected in both replicon vaccinated , as well as live MP12 immunized mice ( Fig . 3 ) , indicating that both the replicon and the live attenuated vaccine elicit a mixed T-helper response . Even though both MP12 infection and the WIV vaccination elicited the highest anti-Gn titers , only the live MP12 infection elicited strong neutralizing antibody responses ( Fig . 4 ) . Gn-C3d-DNA and Gn-C3d-DNA/Rep-Gn vaccinated mice had statistically similar neutralizing titers as MP12 immunized mice . Studies in the past have mainly focused on survival of vaccinated mice post-challenge; however an ideal vaccine should not only be able to protect from virus infection , but also prevent development of clinical symptoms . In this study , we evaluated our candidate vaccines for the ability to confer protection , as well as ability to prevent clinical signs . Few mice from DNA and replicon control groups survived virus infection similar to previous studies [38] , [39] , however all of the control mice displayed clinical signs of infection that was characterized by ruffled fur and lethargy . We observed a correlation between neutralizing antibody titers and development of clinical signs or mortality . Mice with a PRNT50 value of <1∶10 succumbed to lethal infection and a PRNT50 value of ≥1∶40 was sufficient to prevent clinical signs . This indicates that the clinical sickness score more accurately reflected vaccine efficacy in preventing RVFV infection and may be a useful tool for future vaccine studies . The issue of survival from control DNA or mock vaccination is curious , but has been observed in previous publications . Spik et al . [39] also saw survival of a subset of mice following vaccination with DNA controls up to 31 days following challenge with Rift Valley fever virus . In addition , Bird et al . observed that sham mice did not succumb to lethal Rift Valley fever virus challenge , but they developed severe clinical signs of ruffled fur , hunched back , and lethargy [48] . To further explore the ability of factors in the sera to protection mice from RVFV infection , passive transfer of serum from vaccinated mice to naïve mice demonstrated that humoral immune response play a major role in anti-RVFV immunity ( Table 4 ) [40] . Not all mice passively administered the serum were protected , which may be due to dilution of the neutralizing antibodies during preparation that resulted in lack of protection in some mice . Mice vaccinated with replicons alone or in a DNA prime/replicon boost strategy , but not by DNA alone , had robust cellular responses directed at Gn . Cellular responses are critical for clearing virally infected cells in many systems . Although the elicitation of robust neutralizing antibodies are considered ideal for the development of an effective RVFV vaccine , induction of cellular responses by immunization may clear virally infected cells , reduce morbidity , and hasten recovery from infection . The replicon-based vaccines elicited cellular immune responses against the Gn protein , but Gn expressed from DNA plasmids did not , even though priming mice with DNA did not dampen the induction of cellular responses by the Gn-C3d/Rep-Gn in the DNA prime/replicon boost regimen ( Table 2 ) . Although non-specific induction of T-cell responses against RVFV glycoproteins and nucleocapsid proteins have been previously reported [55] , this is the first report to identify an MHC-I restricted immunodominant epitope ( SYAHHRTLL ) on the surface of Gn as predicted by mutlitple algorithm methods to detect the peptide sequence with lowest IC50 and hence better binding to MHC-I [www . immuneepitope . org] . Although , both Rep-Gn alone or in a DNA prime/replicon boost approach Gn-C3d/Rep-Gn induced a combination of humoral and cell mediated immune responses and therefore this strategy may warrant further evaluation in large animals and humans .
|
Rift Valley fever virus ( RVFV ) is an arthropod-borne phlebovirus associated with abortion storms , neonatal mortality in livestock and hemorrhagic fever or fatal encephalitis in a proportion of infected humans . Requirement of multiple booster immunizations to maintain the level of protective immunity with the inactivated vaccines and the ability of live-attenuated vaccines to cause detrimental side-effects are major limitations preventing the widespread use of current vaccines . In this paper , we describe the use of DNA and alphavirus replicon based vaccination approaches to elicit a protective immune response against RVFV . While both vaccines elicited high titer antibodies , DNA vaccination elicited high titer neutralizing antibodies , whereas the replicon vaccine elicited cellular immune responses . Both strategies alone or in combination elicited immune response that completely protected against not only mortality , but also illness . Even though the delivery vectors elicited some protection on their own , they did not prevent severe morbidity . These promising vaccines provide an alternative RVFV vaccine for livestock and humans .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"virology/vaccines",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"immunology/immune",
"response",
"infectious",
"diseases/viral",
"infections"
] |
2010
|
Vaccination with DNA Plasmids Expressing Gn Coupled to C3d or Alphavirus Replicons Expressing Gn Protects Mice against Rift Valley Fever Virus
|
Long-term homologous and temporary heterologous protection from dengue virus ( DENV ) infection may be mediated by neutralizing antibodies . However , neutralizing antibody titers ( NTs ) have not been clearly associated with protection from infection . Data from two geographic cluster studies conducted in Kamphaeng Phet , Thailand were used for this analysis . In the first study ( 2004–2007 ) , cluster investigations of 100-meter radius were triggered by DENV-infected index cases from a concurrent prospective cohort . Subjects between 6 months and 15 years old were evaluated for DENV infection at days 0 and 15 by DENV PCR and IgM ELISA . In the second study ( 2009–2012 ) , clusters of 200-meter radius were triggered by DENV-infected index cases admitted to the provincial hospital . Subjects of any age ≥6 months were evaluated for DENV infection at days 0 and 14 . In both studies , subjects who were DENV PCR positive at day 14/15 were considered to have been “susceptible” on day 0 . Comparison subjects from houses in which someone had documented DENV infection , but the subject remained DENV negative at days 0 and 14/15 , were considered “non-susceptible . ” Day 0 samples were presumed to be from just before virus exposure , and underwent plaque reduction neutralization testing ( PRNT ) . Seventeen “susceptible” ( six DENV-1 , five DENV-2 , and six DENV-4 ) , and 32 “non-susceptible” ( 13 exposed to DENV-1 , 10 DENV-2 , and 9 DENV-4 ) subjects were evaluated . Comparing subjects exposed to the same serotype , receiver operating characteristic ( ROC ) curves identified homotypic PRNT titers of 11 , 323 and 16 for DENV-1 , -2 and -4 , respectively , to differentiate “susceptible” from “non-susceptible” subjects . PRNT titers were associated with protection from infection by DENV-1 , -2 and -4 . Protective NTs appeared to be serotype-dependent and may be higher for DENV-2 than other serotypes . These findings are relevant for both dengue epidemiology studies and vaccine development efforts .
Dengue is caused by four closely related , but antigenically distinct dengue virus serotypes ( DENV-1 , -2 , -3 , -4 ) from the genus Flavivirus in the family Flaviviridae [1] , [2] . In recent decades , dengue has expanded in tropical and subtropical regions and become one of the most prevalent vector-borne diseases of humans with approximately 2 . 5 billion people living with risk of infection . The annual global burden of dengue has been estimated to be 390 million infections with 96 million symptomatic cases [3] . Recently , the first human efficacy trial of a dengue vaccine candidate was completed in Thailand showing good neutralizing antibody response to all four DENV serotypes after vaccination , but no clinical efficacy against DENV-2 infection [4] . The relevance of DENV neutralizing antibodies for protection or modulation of DENV infection , therefore , remains unclear . A primary infection with one serotype is thought to produce long-term protective immunity to re-infection with the homologous serotype . After a limited period of cross-protection , individuals who have had a primary DENV infection are susceptible to infection and disease by heterologous serotypes [5] , [6] . In human challenge studies conducted by Sabin [7] in which DENV naïve individuals were infected with DENV-1 or DENV-2 and re-challenged with homologous or heterologous virus at different times after the initial challenge , protection against disease was observed for at least 18 months against the homologous serotype and at least 2 months against the heterologous serotype . Sabin also noted that infection appeared to be milder if heterologous re-challenge was performed up to 9 months after initial infection suggesting a period of partial heterologous protection . Epidemiological studies of dengue in endemic countries are consistent with this pattern of susceptibility [8] . Mathematical modeling of 38 years of dengue cases admitted to a pediatric hospital in Bangkok , Thailand , was consistent with approximately two years of heterologous protection against disease [9] . Analyses of symptomatic and subclinical DENV infections from prospective cohorts in Thailand and Nicaragua suggested a similar duration of cross-protection against disease [10] , [11] . Heterologous protection is thought to be at least partly mediated by temporary cross-protective neutralizing antibodies from earlier infections [12] . Some studies have shown an association between pre-existing neutralizing antibody titers ( NTs ) and subsequent disease severity under certain conditions [13] , [14] , [15] . Endy et al . found such a correlation between homotypic NTs and subsequent viremia levels and disease severity for DENV-3 , but not for DENV-1 and DENV-2 in a Thai pediatric cohort [14] . In contrast , Sirivichayakul et al . found no relationship between homotypic NTs and subsequent infection by DENV-1 or DENV-4 [16] . Up to now , no epidemiological study in humans has been able to demonstrate an association between pre-existing NTs and protection from infection . One limitation of earlier prospective cohort studies has been that they measured neutralizing antibodies up to one year prior to infection . Neutralizing antibodies ( and especially cross-reactive antibodies ) decrease substantially over time , however , and their kinetics can be quite variable depending on factors such as DENV serotype from previous and current infection , disease severity , host genetics and immunological status [17] . Because neutralizing antibody status just before virus exposure is likely the most relevant for protection from infection , we sought to test the hypothesis that neutralizing antibody titers immediately before exposure was associated with the probability of infection by utilizing data from geographic cluster studies in which high DENV transmission activity has been demonstrated [18] . We showed an association between homotypic NTs and the likelihood of subsequent infection with DENV-1 , -2 and -4 .
Data from two different geographic cluster studies were used in the current analysis . The first study ( called “KPSII” ) was approved by the Institutional Review Boards ( IRBs ) of the Thai Ministry of Public Health ( MOPH ) , Walter Reed Army Institute of Research ( WRAIR ) , University of Massachusetts Medical School ( UMMS ) , University of California at Davis ( UCD ) , and San Diego State University ( SDSU ) . The second study ( called “DEVOL” ) was approved by the IRBs of the Thai MOPH , WRAIR , UCD , and the State University of New York ( SUNY ) Upstate Medical University . Written informed consent was obtained from adult subjects ( age ≥18 years ) or the parents/guardians of child subjects ( age <18 years ) ; assent was obtained from child subjects ≥7 and <18 years of age . In the current study , we used data from a prospective longitudinal cohort and geographic cluster study conducted from 2004 to 2007 among children living in Muang district , Kamphaeng Phet province ( KPP ) in north-central Thailand . The study methodology has been described previously [18] , [19] , [20] . Briefly , geographic cluster investigations were initiated by “index” cases from a longitudinal cohort of approximately 2 , 000 primary school children . Active school absence-based surveillance was used to detect symptomatic DENV infections in the cohort from June to November of each study year . Cohort children who were DENV positive by hemi-nested reverse transcriptase polymerase chain reaction ( PCR ) [21] , [22] from an acute serum sample drawn within three days of illness onset served as an index case to initiate a positive cluster investigation around the index case house . Cohort children who were DENV PCR negative from an acute illness sample served as an index case for a negative cluster investigation . In each geographic cluster , ten to 25 contact subjects aged six months to 15 years living within 100 meters of the index case were enrolled regardless of clinical status . Contact subjects were evaluated at days 0 ( i . e . , same day as cluster initiation ) , 5 , 10 , and 15 by temperature measurement and symptom questionnaire . Serum samples were collected on days 0 and 15 . Paired day 0 and 15 samples underwent DENV nested PCR and an in-house DENV/Japanese encephalitis virus ( JEV ) IgM capture enzyme-linked immunosorbent assay ( ELISA ) [23] . We also used data from a geographic cluster study conducted from 2009 to 2012 in the same district of KPP , Thailand . The study methodology is being submitted in a separate manuscript . In the DEVOL study , geographic cluster investigations were initiated by “index” cases admitted to KPP hospital who were DENV positive by nested PCR [21] , [22] . In each geographic cluster , adults and children ≥6 months of age living within 200 meters of the index case were enrolled regardless of clinical status . The distance was increased compared to the KPSII study to attempt to capture more DENV infections . Only individuals from households where at least one member had a history of fever in the previous seven days were considered for enrollment . Contact subjects were evaluated at days 0 ( i . e . , same day as cluster initiation ) and 14 by temperature measurement , symptom questionnaire , and blood collection . Day 0 samples were tested by DENV nested PCR; day 14 samples were tested using DENV Detect NS1 ELISA ( InBios , Seattle , Washington , USA ) and , if positive , confirmed by DENV nested PCR . Paired day 0 and 14 samples were tested by an in-house DENV/JEV IgM capture ELISA [23] . In order to assess neutralizing antibody status just prior to virus exposure , we first identified contact subjects from the two geographic cluster studies who were DENV positive by nested PCR at day 15 ( KPSII study ) or day 14 ( DEVOL study ) . In the DEVOL study , only clusters in which there was a DENV-2 index case were utilized so that we could obtain data about DENV-2 which had not been detected at day 15 in the earlier KPSII study . Day 14/15 PCR positive subjects were classified as “susceptible” to DENV infection since they had confirmed infection . In order to create a “non-susceptible” comparison group , we identified contact subjects from the two studies who lived in the same house as another subject who was DENV positive on day 0 , but were themselves DENV negative by nested PCR and IgM ELISA on both days 0 and 15 ( KPSII study ) , or by NS1 ELISA and IgM ELISA on both days 0 and 14 ( DEVOL study ) . These DENV negative “non-susceptible” subjects were presumed to have a high likelihood of exposure to DENV during the cluster investigation , yet still did not become infected . In KPSII , contact subjects who met these criteria were selected to be in the “non-susceptible” group . In DEVOL , given the wider age range , each DENV-2 positive “susceptible” subject was matched with two “non-susceptible” subjects by age ( +/− 5 years ) , village and study year . In both the KPSII and DEVOL studies , the exposure serotype for all “non-susceptible” subjects was presumed to be the same as the infecting serotype for the index case from the same cluster . In all subjects , the day 0 blood sample was presumed to reflect the neutralizing antibody status just prior to exposure or infection . The number of DENV-3 infections from both the KPSII and DEVOL studies was not sufficient to be evaluated . To determine neutralizing antibody status just prior to exposure for all “susceptible” and “non-susceptible” subjects , day 0 blood samples were tested by an in-house plaque reduction neutralization test ( PRNT ) using all four DENV serotypes and JEV as previously described [24] , [25] . A monolayer of Macaca mulatta kidney cells ( LLC-MK2 ) was infected with 30–50 plaque-forming units of DENV in the presence of four-fold serial dilutions of heat-inactivated sample on a 12-well plate . For each dilution , the number of virus plaques was counted and compared to the number of plaques in a control where no sample was added . Reference strains were as follows: DENV-1 ( Thailand/16007/1964 ) , DENV-2 ( Thailand/16681/1984 ) , DENV-3 ( Philippines/16562/1964 ) , DENV-4 ( Indonesia/1036/1976 ) [KPSII samples] , DENV-4 ( Thailand/C0036/2006 ) [DEVOL samples] , and JEV ( SA-14-14 vaccine strain ) . In addition to the reference strains , day 0 samples from subjects from DENV-2 clusters ( i . e . , clusters where the index case was infected with DENV-2 ) underwent PRNT using DENV-2 strains from Thailand that were isolated at the AFRIMS laboratory over a period of several decades as follows: Asian-American 1982 ( #D82-165 ) , Asian I 1974 ( #D74-066 ) , Asian I 1984 ( #D84-501 ) , Asian I 1994 ( #D94-035 ) , Asian I 2004 ( #KDS00305 ) , and homologous virus . Homologous viruses were all Asian I genotype and were cultured either from the same blood sample or from a subject from the same cluster or village in the same year . This was done in order to evaluate possible differences in PRNT titers against different DENV-2 strains given the lack of vaccine efficacy against DENV-2 reported in a recent dengue vaccine trial [4] . PRNT data was expressed as the reciprocal of the dilution causing 50% plaque reduction ( PRNT50 ) as extrapolated from probit regression . In this study , our use of the terms “homotypic” and “heterotypic” neutralizing antibodies refers to relationships between serotype neutralizing antibodies based solely on the serotype-specific PRNT results along with the presumptive serotype circulating within a geographic cluster . Analyses were performed comparing all subjects combined ( i . e . , “susceptible” versus “non-susceptible” subjects ) , and by comparing subjects according to exposure serotype . When comparing by serotype , subjects with the same serotype came exclusively from KPSII or DEVOL , but not both . Logistic regression models were employed to test the hypothesis that pre-existing NTs were associated with the probability of being PCR positive or negative . For each test , the dependent variable was the infection status of the subject . For each evaluated DENV strain used for PRNT , a simple model with only the log of NT , and a model adjusted for age were estimated . The model fit and predictive power were assessed for each model using the Akaike Information Criterion ( AIC ) and the area under receiver operating characteristic ( ROC ) curve ( AUC ) . The AIC provides a measure of how closely the models fit the observed data , while penalizing for additional model complexity [26] . Lower AIC values indicate better model fit . The ROC curve is used to evaluate the accuracy of a diagnostic measure [27] . The 95% confidence interval ( CI ) of AUC was computed using 10 , 000 stratified bootstrap replicates [28] . Contingency tables were created to illustrate the predictive ability of individual NT cutoff values . The observed odds ratios were calculated and conditional maximum likelihood estimates of the 95% CIs based on Fisher's exact test were estimated to show the uncertainty associated with the observed odds ratios . All analyses were performed using the R environment for statistical computing . ROC curves were created and AUC was calculated using the pROC package for R [29] .
Of 1599 contact subjects from 50 positive and 53 negative geographic cluster investigations in the KPSII study [18] , [30] , 12 subjects were found to be DENV PCR positive at day 15 and , therefore , considered to be “susceptible”: six had DENV-1 and six had DENV-4 . All 12 subjects were PCR negative at day 0 , and DENV IgM ELISA negative at days 0 and 15 . No data or blood was collected after day 15 . Twenty-two subjects from KPSII were selected as “non-susceptible”: 13 from DENV-1 clusters and nine from DENV-4 clusters . The predominant circulating serotype during the first two years of the KPSII study was DENV-4 and the last two years was DENV-1 [19] . Of 740 contact subjects from 195 DENV-2 geographic cluster investigations in the DEVOL study , five subjects were found to be DENV-2 PCR positive at day 14 and , therefore , considered as “susceptible . ” All five subjects were PCR negative at day 0 , and DENV IgM ELISA negative at days 0 and 14 . No data or blood was collected after day 14 . Ten subjects were selected as “non-susceptible” from DENV-2 clusters . The predominant circulating serotype during all years of the DEVOL study was DENV-2 ( unpublished data ) . Altogether , 49 subjects were available for analysis: 17 PCR positive “susceptible” subjects ( six DENV-1 , five DENV-2 and six DENV-4 ) , and 32 PCR negative “non-susceptible” subjects ( 13 from DENV-1 clusters , 10 DENV-2 and 9 DENV-4 ) . Table 1 lists characteristics of the 49 subjects . Figures 1 , 2 and 3 show bar graphs of day 0 NTs for each subject . Seven “susceptible” compared with three “non-susceptible” subjects had DENV naïve NT profiles on day 0 ( odds ratio = 6 . 460 [95% CI 1 . 198 , 46 . 302] ) . When all serotypes were combined , PCR status was significantly associated with the log of homotypic NT both alone and adjusted for age ( Table 2 ) . ROC curves were created for all serotypes combined , and for each serotype separately ( Figure 4 ) . Considering subjects from just DENV-1 clusters , age-adjusted models to predict PCR status were better ( based on AIC and AUC ) when using homotypic DENV-1 NTs ( AUC = 0 . 833 [95% CI 0 . 590 , 1 . 000] ) than heterotypic NTs ( Table 3 ) . For DENV-4 clusters , age-adjusted models were better using homotypic DENV-4 NTs ( AUC = 0 . 889 [95% CI 0 . 667 , 1 . 000] ) than heterotypic NTs ( Table 3 ) . For DENV-2 clusters , age-adjusted models were better using both homotypic DENV-2 reference NTs ( AUC = 0 . 740 [95% CI 0 . 460 , 0 . 960] ) and heterotypic DENV-4 reference NTs ( AUC = 0 . 880 [95% CI 0 . 680 , 1 . 000] ) than other heterotypic NTs ( Table 3 ) . When different DENV-2 strains were used for PRNT , the AUC using Asian I 1974 strain appeared to have the best fit . The AUC using DENV-2 homologous virus was comparable to that using DENV-2 reference strain . Homotypic reference strain NT cutoff values as determined by ROC curves were used to create two-by-two contingency tables to demonstrate the relationship between individual NT cutoffs and PCR status ( Table 4 ) . Homotypic NT cutoff values were 11 , 11 , 16 and 323 for all serotypes combined , DENV-1 , DENV-4 , and DENV-2 , respectively . The observed odds ratio for each sample is indicated and the 95% CI is presented to show the uncertainty associated with this measure . These CIs should not be used to assess statistical significance because they are post-hoc comparisons based on the cutoffs indicated by the ROC plot analysis . For all serotypes combined , the observed odds ratio of becoming infected ( i . e . , being PCR positive ) if day 0 homotypic NT was ≥11 was 0 . 153 ( 95% CI 0 . 023 , 0 . 700 ) . For DENV-1 clusters , the observed odds ratio if homotypic NT was ≥11 was 0 ( 95% CI 0 . 000 , 1 . 554 ) . For DENV-4 clusters , the observed odds ratio if homotypic NT was ≥16 was 0 ( 95% CI 0 . 000 , 0 . 842 ) . For DENV-2 , the odds ratio if homotypic NT was ≥323 was 0 ( 95% CI 0 . 000 , 1 . 279 ) .
Our analysis of subjects from geographic cluster studies indicates that pre-existing homotypic neutralizing antibody titers as measured by PRNT were positively associated with protection against infection by DENV-1 , -2 and -4; a similar analysis could not be performed for DENV-3 because too few cases were available . Homotypic NTs were more strongly associated with protection than heterotypic NTs except in the case of DENV-2 infections , in which pre-existing heterotypic DENV-4 NTs were also positively associated with protection against DENV-2 . This is the first human epidemiological study to report such an association and provides some support for the use of PRNT titers as a correlate of protective immunity in dengue epidemiology studies and vaccine development efforts . Two major strengths of this study design likely explain why a significant association between NTs and protection from infection was observed in contrast to previous prospective cohort studies . Whereas earlier cohort studies typically analyzed NTs in blood samples collected up to six months or more prior to the incident DENV infection , our study provided blood samples from within two weeks prior to virus exposure , thus minimizing the confounding factor of variations in antibody kinetics . Although we could not document the exact day of virus inoculation in “susceptible” subjects , the two week interval was likely sufficient to ensure that day 0 blood collections occurred before exposure . This presumption is further supported by the fact that DENV IgM ELISA was negative at day 14/15 indicating that infection was still early in its course . Although we were able to confirm infection in “susceptible” subjects by a positive PCR , we could not prove exposure in PCR negative subjects . By requiring documentation of a DENV infection in the same house , however , our selection criteria for “non-susceptible” subjects maximized the likelihood of such exposure . Previous results from our cluster studies indicate a very high likelihood of virus exposure in houses with known DENV infection [18] . We have reported average DENV infection rates of greater than 30% over two weeks in houses with known DENV infection pointing to even higher rates of virus exposure if individuals who are exposed but immunologically protected are taken into account . These exposure rates are as high as is reasonably feasible in the natural setting ( i . e . , outside of an experimental challenge ) . Whereas relatively low homotypic NT cutoff values were associated with protection against DENV-1 and DENV-4 in this study , much higher homotypic NT cutoffs were associated with protection against DENV-2 . This indicates that protective NTs may depend , in part , on infecting serotype , with DENV-2 possibly requiring substantially higher PRNT titers . This finding needs to be tempered by the uncertainty in the mix of functional and non-functional antibody subpopulations that are being measured by PRNT , which could affect the generalizability of these NT cutoff values to other populations . Furthermore , although age was incorporated into the statistical analyses , the inclusion of adults in the DEVOL study compared to only children in the KPSII study could have introduced an additional confounding factor in interpreting NT values . Individual PRNT titers necessary for protection may , therefore , be difficult to define precisely . This difficulty could be compounded by the inherent variability in biological assays such as PRNT whether using similar or different methods ( e . g . , different cell lines ) [25] . Nevertheless , our findings highlight the likelihood that traditionally accepted NT ranges for vaccine immunogenicity may not necessarily be relevant for protection from a natural DENV challenge . The lack of clinical efficacy against DENV-2 in the recent dengue vaccine trial in Thailand [4] may have partly been due to insufficiently high DENV-2 NTs . Moreover , vaccination in a primed population with the potential for interference from existing immunity makes this situation even more complex . What may be a protective NT level after natural DENV infection may not apply in the setting of multivalent vaccination , especially in DENV primed individuals . Although our study indicates that PRNT titers may be related to protection from infection , these titers are likely to be surrogates for other immune responses that are more directly relevant to protection . For example , PRNT titers measured in Fc gamma receptor-bearing cells [31] , [32] or after depletion of cross-reactive antibodies [33] , and quaternary E-protein domain I/II hinge region antibody titers [34] have been proposed to be better indicators of protection . Conversely , E-protein domain II fusion loop antibodies have been proposed to function as cross-protective antibodies within the pool of heterotypic neutralizing antibodies [35] . High frequencies of circulating DENV-specific T lymphocyte responses have also been proposed to provide some measure of protective immunity [36] , [37] . Whatever measure is used will need to be validated with well characterized samples from studies with clinically relevant outcomes . In our study , it is interesting that PRNT using homologous DENV-2 strains did not strengthen the association with protection as compared with PRNT using DENV-2 reference strain . Furthermore , using an altogether different DENV-2 genotype ( e . g . , Asian-American ) from the infecting Asian I genotype did not weaken the association . These results suggest that strain differences in PRNT may be of little importance in assessing protection . Other more functionally relevant assays may still be able to detect clinically relevant strain-related differences . Although the samples in our study were unique and informative , the number that met testing criteria was small , limiting the power of our analysis . Nevertheless , the fact that significant associations were found despite these small numbers supports the strength of the relationships . Larger sample numbers may have revealed more subtle associations , for example , between different heterotypic NT combinations and protection from infection . Because of the small numbers , we used data from two different geographic cluster studies over multiple years . The inclusion of both adults and children with likely differing immunological backgrounds may have affected our findings for the DENV-2 analysis , possibly accounting for the higher DENV-2 NTs associated with protection . We also found that “susceptible” subjects were more likely to be DENV NT naïve than “non-susceptible” subjects . Thus , part of the increased likelihood of infection associated with lower baseline NTs may have been due to susceptibility to primary versus secondary infections . In addition , as mentioned earlier , we could not be certain that “non-susceptible” subjects had , in fact , been exposed to DENV infection . If , however , some of the “non-susceptible” subjects had simply been unexposed rather than protected from infection , the likelihood of detecting an association between NT and protection would have been even less . Finally , we were not able to characterize the clinical status of infected subjects because no follow up visits took place after day 15 . Therefore , we were unable to make conclusions about associations with clinical severity which may be a more relevant endpoint for vaccine evaluation than detection of viremia by PCR . In our study population , neutralizing antibody titers were associated with protection against DENV-1 , -2 and -4 . NT levels required for this protection varied for these three serotypes , but were likely affected by the preceding epidemiological and immunological history of the subjects . These findings will help inform ongoing studies of dengue epidemiology and the development of dengue vaccine candidates .
|
Dengue is caused by four different dengue virus serotypes ( DENV-1 , -2 , -3 , -4 ) . Infection induces long-term protection against the same serotype , but only short-term protection , and possible enhancement , from different serotypes . DENV neutralizing antibody titers ( NTs ) are thought to mediate protection or modify disease . Association of NTs with protection from infection has not , however , been clearly demonstrated . We analyzed data from two geographic clusters studies conducted in Kamphaeng Phet , Thailand , in which DENV NTs just before virus exposure were compared between DENV-infected “susceptible” and non-infected “non-susceptible” subjects . NTs appeared to be associated with protection against DENV-1 , -2 , and -4 , but at different NT cutoff levels , with the cutoff for DENV-2 appearing to be the highest . These findings are relevant for ongoing efforts to investigate dengue epidemiology and develop dengue vaccine candidates .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"viral",
"hemorrhagic",
"fevers",
"infectious",
"diseases",
"infectious",
"disease",
"immunology",
"medicine",
"and",
"health",
"sciences",
"clinical",
"immunology",
"dengue",
"fever",
"biology",
"and",
"life",
"sciences",
"immunology",
"viral",
"diseases",
"vector-borne",
"diseases"
] |
2014
|
Dengue Virus Neutralizing Antibody Levels Associated with Protection from Infection in Thai Cluster Studies
|
Gonadal failure , along with early pregnancy loss and perinatal death , may be an important filter that limits the propagation of harmful mutations in the human population . We hypothesized that men with spermatogenic impairment , a disease with unknown genetic architecture and a common cause of male infertility , are enriched for rare deleterious mutations compared to men with normal spermatogenesis . After assaying genomewide SNPs and CNVs in 323 Caucasian men with idiopathic spermatogenic impairment and more than 1 , 100 controls , we estimate that each rare autosomal deletion detected in our study multiplicatively changes a man's risk of disease by 10% ( OR 1 . 10 [1 . 04–1 . 16] , p<2×10−3 ) , rare X-linked CNVs by 29% , ( OR 1 . 29 [1 . 11–1 . 50] , p<1×10−3 ) , and rare Y-linked duplications by 88% ( OR 1 . 88 [1 . 13–3 . 13] , p<0 . 03 ) . By contrasting the properties of our case-specific CNVs with those of CNV callsets from cases of autism , schizophrenia , bipolar disorder , and intellectual disability , we propose that the CNV burden in spermatogenic impairment is distinct from the burden of large , dominant mutations described for neurodevelopmental disorders . We identified two patients with deletions of DMRT1 , a gene on chromosome 9p24 . 3 orthologous to the putative sex determination locus of the avian ZW chromosome system . In an independent sample of Han Chinese men , we identified 3 more DMRT1 deletions in 979 cases of idiopathic azoospermia and none in 1 , 734 controls , and found none in an additional 4 , 519 controls from public databases . The combined results indicate that DMRT1 loss-of-function mutations are a risk factor and potential genetic cause of human spermatogenic failure ( frequency of 0 . 38% in 1306 cases and 0% in 7 , 754 controls , p = 6 . 2×10−5 ) . Our study identifies other recurrent CNVs as potential causes of idiopathic azoospermia and generates hypotheses for directing future studies on the genetic basis of male infertility and IVF outcomes .
Male infertility is a multifaceted disorder affecting nearly 5% of men of reproductive age . In spite of its prevalence and a considerable research effort over the past several decades , the underlying cause of male infertility is uncharacterized in up to half of all cases [1] . Some degree of spermatogenic impairment is present for most male infertility patients , and , in its most severe form , manifests as azoospermia , the lack of detectable spermatozoa in semen , or oligozoospermia , defined by the World Health Organization as less than 15 million sperm/mL of semen . Spermatogenesis is a complex multistep process that requires germ cells to ( a ) maintain a stable progenitor population through frequent mitotic divisions , ( b ) reduce ploidy of the spermatogonial progenitors from diploid to haploid through meiotic divisions , and ( c ) assume highly specialized sperm morphology and function through spermiogenesis . These steps involve the expression of thousands of genes and carefully orchestrated interactions between germ cells and somatic cells within the seminiferous tubules [2] . It is likely that a large proportion of idiopathic cases of spermatogenic failure are of uncharacterized genetic origin , but measuring the heritability of infertility phenotypes has been challenging . Known genetic causes of non-obstructive azoospermia ( NOA ) include deletions in the azoospermia factor ( AZF ) regions of the Y chromosome [3] , Klinefelter's syndrome [4] , and other cytogenetically visible chromosome aneuploidies and translocations [5] . Beyond these well-established causes , which are observed in 25–30% of cases , the genetic architecture of spermatogenic impairment is currently unknown . One might expect a priori that rare or de novo , large effect mutations will be the central players in genetic infertility , and indeed other primary infertility phenotypes like disorders of gonadal development , isolated gonadotropin-releasing hormone deficiency , and globozoospermia , a disorder of sperm morphology and function , appear to be caused by essentially Mendelian mutations operating in a monogenic or oligogenic fashion [6] , [7] , [8] . Similarly , recurrent mutations of the AZF region on the Y chromosome are either completely penetrant ( AZFa , AZFb/c ) or highly penetrant ( AZFc ) risk factors for azoospermia . Our working model at the start of this study was that additional “AZF-like” loci existed in the genome , either on the Y chromosome or elsewhere , and that , much like recent progress in the analysis of developmental disorders of childhood , a large number of causal point mutations and submicroscopic deletions could be revealed in idiopathic cases by the appropriate use of genomic technology . In this paper , we employ oligonucleotide SNP arrays as discovery technology to conduct a whole-genome screen for two rare genetic features in men with spermatogenic failure . First , we extract and analyze the probe intensity data to find rare copy number variants ( CNVs ) . A growing number of CNVs have been associated with a host of complex disease states [9] including neurological disorders [10] , [11] , [12] , [13] , several autoimmune diseases [14] , [15] , type 2 diabetes [16] , cardiovascular disease [17] , and cancer [18] , [19] , [20] , [21] . Now , a role for CNVs in male infertility is beginning to emerge [22] , [23] , [24] , [25] . As a second approach to identify rare genetic variants , we use a population genetics modeling framework to identify large homozygous-by-descent ( HBD ) chromosome segments that may harbor recessive disease alleles . When applied to consanguineous families , so-called “HBD-mapping” has been an unequivocal success in identifying the location of causal variants for simple recessive monogenic diseases [26] . HBD analysis can also be used to screen for the location of rare variants in common disease case-control studies of unrelated individuals , using either a single-locus association testing framework or by testing for an autozygosity burden , frequently referred to as “inbreeding depression”: an enrichment of size or predicted functional impact of HBD regions aggregated across the genome . This approach has produced results for a growing list of common diseases , including schizophrenia [27] , Alzheimer's disease [28] , breast and prostate cancer [29] . In this study , we screened three cohorts of men with idiopathic spermatogenic failure in an attempt to identify rare , potentially causal mutations , and to better understand the genetic architecture of the disease ( Table 1 ) . We found a genomewide enrichment of large , rare CNVs in men with spermatogenic failure compared to normozoospermic or unphenotyped men ( controls ) . We also identify a number of cases with unusual patterns of homozygosity , possibly the result of recent consanguineous matings . Our results show that spermatogenic output is a phenotype of the entire genome , not just the Y chromosome , place spermatogenic failure firmly among the list of diseases that feature a genomewide burden of rare deleterious mutations and provide a powerful organizing principle for understanding male infertility .
When restricting our analysis to CNVs with a call frequency of less than 5% , a subset likely to be enriched for pathogenic events , we observed pronounced differences among groups ( Table S1 ) . Azoospermic and oligozoospermic men have nearly twice the amount of deleted sequence genomewide when compared to controls ( p = 1 . 7×10−4 , Wilcoxon rank sum test ) , and a nonsignificant 12% increase in the number of deletions per genome . When examining the even more restricted set of rare CNVs larger than 100 kb ( Dataset S1 ) , these associations are more pronounced: the rate of deletions in cases was twice that of controls ( 1 . 12 vs . 0 . 55 , p = 9 . 7×10−4 ) and the amount of deleted sequence 2 . 6 times greater in cases ( p = 8 . 8×10−4 ) . In order to replicate these initial findings , we assayed two additional cohorts – one group of 61 Caucasian men with severe spermatogenic impairment and 100 ethnicity-matched , unphenotyped controls , both collected at Washington University in St . Louis ( WUSTL ) , and a larger case cohort of 179 Caucasian men with idiopathic azoospermia , primarily from medical practices in Porto , Portugal , matched to an unphenotyped control set of 974 Caucasian men collected by the UK National Blood Service ( NBS , [30] ) . Although using different array platforms ( Text S1 ) , we observed replication of our initial association ( Table S2 and Table S3 ) ; in the WUSTL cohort a 20% increase in the rate ( p<0 . 05 ) and in the Porto cohort a 31% increase in rate ( p<5×10−3 ) . We excluded several artifactual explanations for this burden effect , including specific batch phenomena or population structure ( Text S1 , Figures S1 , S2 , S3 , S4 , S5 ) . To better characterize these genomewide signals , we set out to search for clustering of pathogenic mutations on specific chromosomes . We focused first on the Y chromosome as it is the location of most known mutations modulating human spermatogenesis ( Figure 1 , Figure S6 ) . Y-linked microdeletions of the AZFa , AZFb , and AZFc regions are well-established causes of spermatogenic impairment , and thus we excluded from this study cases with AZF microdeletions visible by STS PCR . In the array data , we found no significant difference in the frequency of rare Y deletions between case and controls groups; however rare duplications were more abundant in Porto cases compared to the NBS controls ( a 3-fold enrichment in Porto cohort , p = 1 . 9×10−3 ) . We could classify the majority ( >90% ) of our samples to major Y haplogroups using SNP genotypes ( Text S1 ) , and , as expected , most of these samples fall into the two most common European haplogroups: I ( 22% ) and R ( 70% ) . The observed duplication burden was not an artifact of differences in major Y haplogroup frequency between cases and controls , as association was essentially unchanged when only considering samples with haplogroup R1 ( p = 3 . 3×10−3 ) . Due to low probe coverage , only one Y-linked duplication was called in the Utah cohorts ( in a control individual ) and two in the WUSTL cohort ( both in cases ) , so this burden of Y duplications was not replicated . Next we turned to the X chromosome , which is highly enriched for genes transcribed in spermatogonia [31] . In the Utah cohorts there were 71 gains and losses with a frequency of less than 5% on the X chromosome , cumulatively producing three times as much aneuploid sequence in azoospermic and oligozoospermic men compared with normozoospermic men ( 89 kb/person azoo , 45 kb/person oligo , 27 kb/person normozoospermic men , all cases versus controls p<0 . 03 ) . This burden was strongly replicated in the Porto samples , which displayed a 1 . 6 fold enrichment of rare CNV on the X ( p = 5×10−4 ) and the WUSTL samples ( 31% of cases with a rare X-linked CNV versus 16% of controls , p = 0 . 02 by permutation ) . The genome-wide signal of CNV burden was not driven solely by sex chromosome events: considering only autosomal mutations in Utah samples there was an enrichment of aneuploid sequence in large deletions in azoospermic men ( 268 kb/person ) and oligozoospermic men ( 308 kb/person ) compared to control men ( 189 kb/person , p = 9 . 8×10−3 ) , and an enrichment in the rate of deletions in all cases when considering just events >100 kb ( 1 . 9 fold enrichment , p = 6×10−3 ) . In the Porto cohort , there was modest evidence for a higher rate of rare deletions of all sizes in azoospermic men ( 1 . 27 fold enrichment , not significant ) as well as an increase in total amount of deleted sequence ( 345 kb/case vs . 258 kb/control , p<0 . 003 ) . In order to cleanly summarize our findings across all cohorts , we fit logistic regression models for each cohort , regressing case status onto CNV count for different classes of CNV . We also fit a linear mixed-effects logistic regression model to the total dataset for each CNV class , treating cohort as a random factor ( Figure 1 ) . In each regression model we controlled for population structure by including eigenvectors from a genomewide principal components analysis ( Methods ) . On the basis of the combined analysis , we estimate that each rare autosomal deletion multiplicatively changes the odds of spermatogenic impairment by 10% ( OR 1 . 10 [1 . 04–1 . 16] , p<2×10−3 ) , each rare X-linked CNV ( gain or loss ) by 29% , ( OR 1 . 29 [1 . 11–1 . 50] , p<1×10−3 ) and each rare Y-linked duplication by 88% ( OR 1 . 88 [1 . 13–3 . 13] , p<0 . 03 ) . Deletions of the AZF regions of the Y chromosome are often mediated by non-allelic homologous recombination ( NAHR ) between segmental duplications and are the most common known cause of spermatogenic failure . Because of their prognostic power and high rate of recurrence in the population , screening for AZF deletions is a standard part of the clinical workup for azoospermia . It would be of high clinical value if additional azoospermia susceptibility loci with significant recurrence rates could be identified . We screened all cohorts for large ( >100 kb ) rearrangements flanked by homologous segmental duplications capable of generating recurrent events by NAHR [32] . There was no significant enrichment of gains or losses in cases across these hotspot regions when considered as an aggregate . Due to small sample sizes we found no single-locus associations , at these hotspot loci , or elsewhere , that met the strict criteria of genomewide significance in both the discovery and replication cohorts . Many of our single-cohort associations from one platform lack adequate probe coverage on other platforms for robust replication ( Text S1 ) . However , several loci were significant on joint analysis of all cohorts . The best candidate for a novel locus generating NAHR-mediated infertility risk mutations is a 100 kb segment on chromosome Xp11 . 23 flanked by two nearly identical ( >99 . 5% homology ) 16 kb segmental duplications containing the sperm acrosome gene SPACA5 ( Figure 2a , Figure S7 ) . We identified 9 deletions of this locus spread across all patient cohorts ( 3 in PT , 1 in UT , 5 in WUSTL ) compared to 8 in the pooled 1124 controls ( 2 . 8% frequency versus 0 . 7% , odds ratio = 3 . 96 , p = 0 . 005 , Fisher exact test ) . We genotyped the deletion by +/− PCR in an additional cohort of 403 men with idiopathic NOA from Weill Cornell , and observed an additional 3 deletions ( Figure S8 , Text S1 ) . In a prior case-control study of intellectual disability , investigators using qPCR estimated the allele frequency of this deletion to be 0 . 47% ( 10/2121 ) in a large Caucasian male control cohort [33] . Combining these data , we estimate the allele frequency of the deletion to be 1 . 6% in Caucasian cases , compared to 0 . 55% in Caucasian controls ( OR 3 . 0 , 95% CI 1 . 31–6 . 62 , p = 0 . 007 ) . The deleted region contains the X-linked cancer-testis ( CT-X ) antigen gene SSX6; the CT-X antigen family is a highly duplicated gene family on the X chromosome comprising 10% of all X-linked genes and is expressed specifically in testis . After controlling for differences in coverage across the array platforms used in this study , we find a significant enrichment of rare deletions of CT-X genes in all cases ( p = 0 . 02 ) ; this finding did not extend to duplications or CT antigen genes on the autosomes ( Table 2 ) . When analyzing all cohorts jointly , our strongest association ( genomewide corrected p-value <0 . 002 ) is to both gains and losses involving a 200 kb tandem repeat on Yq11 . 22 , DYZ19 ( Figure S6 , Figure S9 ) , a human-specific array of 125 bp repeats first discovered as a novel band of heterochromatin in the Y chromosome sequencing project [34] . Tandem repeat arrays are often highly unstable sequence elements that can mutate by both replication-based and recombination-based ( e . g . NAHR ) mechanisms . In our data there were 9 gains and 11 losses at DYZ19 in 323 cases ( combined frequency 6 . 1% ) , compared to 3 gains and 12 losses in 1136 controls ( combined frequency 1 . 3% ) . While this finding may ultimately require painstaking technical work to conclusively validate , we have several reasons to believe the association is real . First , we have previously shown that it is possible to identify real copy number changes at VNTR loci using short oligonucleotide arrays [35]; second , copy number changes at this locus were identified by multiple platforms in the current study; third , the association is nominally significant in both the Utah and Porto cohorts; fourth the locus is within the AZFb/c region . The direction of copy number changes does appear to track with haplogroup – while 12/13 duplications occur on the R1 background , 14/15 deletions for which haplogroup could be determined occur on I or J background . Haplogroup assignments for the carriers of these CNVs were confirmed by standard short tandem repeat analysis ( Text S1 ) . The strong association between haplogroup and direction of copy number change is noteworthy; it may indicate that DYZ19 CNVs are merely correlated with other functional changes on these chromosomes , or perhaps the structure of these chromosomes predisposes them to recurrent gains ( R1 ) or losses ( I/J ) . The gene DMRT1 is widely believed to be the sex-determination factor in avians , analogous to SRY in therians , and may play the same or similar role in all species that are based upon the ZW sex chromosome system [36] . DMRT1 encodes a transcription factor that can activate or repress target genes in Sertoli cells and premeiotic germ cells through sequence-specific binding [37] . In humans , DMRT1 is located on 9p24 . 3 in a small cluster with the related genes DMRT2 and DMRT3 . Large terminal deletions of 9p are a known cause of syndromic XY sex-reversal , and although the role of the DMRT genes in the 9p deletion syndrome phenotype has not yet been defined , mouse experiments have shown that homozygous deletion of DMRT1 causes severe testicular hypoplasia [38] , [39] , [40] . We found two , perhaps identical , 132 kb deletions spanning DMRT1 in the Utah cohort in men with azoospermia , and a 1 . 8 Mb terminal duplication of 9p , spanning these genes , was seen in a single normozoospermic control from Utah ( Figure 2b ) . All three of these rearrangements were validated by TaqMan assay ( Figure S10 , Text S1 ) . Both men were recruited into the study in Salt Lake City , UT between 2002 and 2004 . They self-reported their ancestry as Caucasian , and in both cases this assumption was clearly verified by principal components analysis of their genetic data ( Figure S2 ) . There was no evidence that the two deletion carriers were closely related upon comparison of their whole-genome SNP genotypes . Testis biopsies were performed on both men; these indicated apparent Sertoli cell only syndrome in the first and spermatocytic arrest in the second . Both men exhibited apparently normal male habitus and virilization with no phenotypic similarities to 9p deletion syndrome . We obtained Affymetrix 6 . 0 array data from a previously published genomewide association study of idiopathic NOA in Han Chinese [41] comprised of 979 cases and 1734 controls ( Text S1 ) . After processing these samples with our CNV calling pipeline , we observed an additional 3 deletions of DMRT1 exonic sequence in cases ( 0 . 3% ) and none in controls ( Figure 2B , Figure S11 ) . From these combined array data we estimate a frequency of DMRT1 exonic deletion of 0 . 38% ( 5/1306 ) in cases and 0% ( 0/2858 ) in controls ( OR = Infinity , [2 . 0-Inf] , p = 0 . 003 ) . We obtained the two largest control SNP array datasets in the Database of Genomic Variants ( DGV ) , representing CNV calls from 4519 samples typed with platforms of equal or higher probe density to the ones used here [42] , [43] . None of these samples contained CNV of any sort affecting DMRT1 . Finally , we screened an additional set of 233 idiopathic NOA cases from Weill Cornell , and 135 controls with the TaqMan validation assay and identified an additional 3 deletions ( 2 in cases , 1 in controls , Text S1 , Figure S12 ) . As this qPCR assay interrogates intronic sequence , the functional consequences of these 3 deletions are unclear . Our array data have revealed some of the smallest coding deletions of DMRT1 reported to date in humans , and should help to clarify the critical regions of 9p involved in testicular development and function . Notably , using a bespoke reanalysis of the intensity data , we did not see evidence for CNVs involving the gene PRDM9 , a recently characterized zinc finger methyltransferase that appears to control the location of recombination hotspots in a diversity of mammalian species . Heterozygosity of PRDM9 zinc finger copy number has been shown to cause sterility in male hybrids of Mus m . domesticus and Mus m . musculus due to meiotic arrest [44] . The identification of functional or physical annotations enriched in case-associated CNVs can be a powerful step in constructing models to classify pathogenic variants . We searched for significant case-specific aggregation of CNVs in several classes of functional sequence , including 195 genes previously shown to result in spermatogenic defects when mutated in the mouse [45] , all protein and non-protein coding genes , and 525 testis genes that are differentially expressed during human spermatogenesis ( Text S1 ) . Deletion of X- or Y-linked exonic sequence conferred the strongest risk ( OR = 1 . 87 [1 . 30–2 . 68] , p<1×10−3 ) . Very similar risk was associated with deletion of exonic sequence from testis genes differentially expressed during spermatogenesis , despite the fact that only 15% of these genes are located on the sex chromosomes ( OR = 1 . 85 [1 . 01–3 . 39] , p<0 . 05 ) . Deletion of any exonic sequence was also associated with disease ( OR = 1 . 25 [1 . 07–1 . 46] , p<5×10−3 ) . Deletion of miRNAs was not associated , nor was deletion of the 195 mouse spermatogenic genes [45] , which were very rarely deleted in either cases or controls . We hypothesized that at least some of the functional impact of CNV burden on fertility was a result of disruption of haploinsufficient ( HI ) genes , as has been demonstrated for neuropsychiatric and developmental disease [46] . For each singleton deletion in our collections we used a recently described modeling framework to calculate the probability that the deletion is pathogenic due to dominant disruption of a haploinsufficient gene [47] . Much to our surprise , HI scores from deletions in infertility cases were much smaller than those from cases of autism and developmental disorders and in fact indistinguishable from controls ( mean HI score −1 . 16 in controls , −1 . 02 in all spermatogenic impairment cases , p = 0 . 49 by Wilcoxon rank sum test; Figure 3 ) . Likewise there was no enrichment of large rearrangements within 45 known genomic disorder regions in cases [46] . In contrast to previously described diseases that feature CNV burden , spermatogenic impairment may be more likely to result from large effect recessive mutations , or perhaps the additive effect of deleterious mutations across many loci . We sought to uncover support for recessive mutation load in our cases by assessing the impact of inbreeding , or elevated rates of homozygosity , on disease risk by applying a population genetic approach to the SNP genotype data from our samples [48] . The major genetic side effect of consanguineous mating is a genome-wide increase in the probability that both paternal and maternal alleles are homozygous-by-descent . This probability is often summarized as the inbreeding coefficient , F , and can be estimated from analysis of pedigree structure or by direct observation of genomewide SNP genotypes . Due to differences in demographic history and culture , the extent of background homozygosity in the genome is expected to vary when comparing diverse populations throughout the globe . The haplotype modeling algorithms implemented in the software package BEAGLE estimate the background patterns of linkage disequilibrium and homozygosity across a set of samples , allowing population-specific information to be used to assess the evidence that any given section of a genome is likely to be homozygous-by-descent ( HBD ) . During the course of our study we concluded that standard PCA-based approaches to stratification are insufficient to correct for population structure during the analysis of inbreeding , even when using population genetic methods like BEAGLE ( Text S1 , Figure S13 ) . The problem comes not from spurious identification of HBD , but from spurious association of HBD with disease status when case and controls are sampled from groups with different levels of background relatedness . For instance , in a recent survey of 17 Caucasian cohorts , estimates of the average inbreeding coefficient , F , varied from 0 . 09% to 0 . 61% , with UK-based cohorts showing the lowest F and the one Portuguese cohort showing the highest [27] . While PCA-based methods traditionally detect and correct for differences in allele frequencies among groups , we believe that they do not detect differences in inbreeding that can be readily incorporated into a case-control testing framework . In the following section , we use data from 622 healthy adults from Spain , who we believe form a more appropriate control group for the Porto case cohort ( Methods , Text S1 , Figure S13 ) . Analyzing each cohort separately , BEAGLE identified 5343 chromosome segments likely to represent HBD regions ( HBDRs ) across all samples . We excluded low-level admixture as a spurious source of HBD ( Figure S3 ) . Only three of these segments were identified as apparent artifacts induced by large heterozygous deletions ( 287 kb , 817 kb , and 877 kb in size ) and were removed before subsequent analyses . As expected , the distribution of HBD across all samples was L-shaped , with the majority of HBDRs shorter than 1 Mb and a few intermediate and very large events observed ( Figure 4b ) . The largest HBDR identified spanned all of chromosome 2 in an azoospermic individual , indicative of uniparental isodisomy of the entire chromosome . Clinical reports of UPD2 are extremely rare – there are 7 previous reports of UPD2 that have been ascertained through association with an autosomal recessive disorder [49] . In each of these cases a recessive disorder that lead to clinical presentation was identified . There is currently no proof of imprinted genes on chromosome 2 from either mouse or human data . We performed whole exome sequencing on this individual , and using a simple scoring scheme based on functional annotation and population genetic data , identified a homozygous missense mutation of the INHBB gene as the most unusual damaging homozygous lesion in the genome of this individual ( Figure 5 , Text S1 ) . The biology of the INHBB gene product strongly implicates this mutation as a causal factor but without additional functional or epidemiological evidence such a conclusion is speculative ( Figure 6 ) . Setting aside this case of UPD2 , we found only modest evidence for an enrichment of homozygosity in men with spermatogenic impairment ( Figure 4a , Table 3 ) . Our hypothesis was that , if a large percentage of cases of azoospermia were attributable to large-effect autosomal recessive Mendelian mutations , we would see a corresponding increase in the proportion of cases with large values of F . The average inbreeding coefficient was numerically higher in each case cohort compared to its matched control cohort ( Table 3 ) . We used a logistic regression mixed model framework to test for association between autozygosity and disease , while controlling for population structure , fitting models that treated autozygosity as both a categorical variable ( e . g . inbreeding coefficient >6 . 25% , yes or no ) and a continuous variable ( F , Methods ) . While the estimated effect of inbreeding on disease risk was positive in every model that we tested , the corresponding odds ratios did not differ significantly from 1 in any version ( Table 3 ) . There were fewer than 10 HBD regions shared by 2 or more cases , supporting the model that spermatogenic efficiency has a polygenic basis . We also tested for case-specific aggregation of HBD segments using the same association framework as that used for CNVs . We did not identify any significant patterns . Based on published analyses of small-effect recessive risk mutations in other complex diseases , we believe our current sample size would be underpowered to detect association between very old inbreeding ( e . g . due to shared ancestors 15 generations ago ) . It is possible that large cohorts , consisting of over 10 , 000 cases , may be needed to accurately estimate the relationship between low-level variation in inbreeding ( F values smaller than 0 . 1 ) and azoospermia risk , as well as map specific risk alleles [27] , [50] .
We report here the largest whole genome study to date investigating the role of rare variants in infertility , examining data from 323 cases of male infertility and 1 , 136 controls . These data demonstrate that rare CNVs are a major risk factor for spermatogenic impairment , and while confirming the central role of the Y chromosome in modulating spermatogenic output , our risk estimates for autosomal and X-linked CNVs indicate that this phenotype is influenced by rare variation across the entire genome . The controls from two of the cohorts were unphenotyped , and given the estimated prevalence of azoospermia ( 1% ) , we may have underestimated the risk associated with these large rearrangements . We observed 5 deletions of DMRT1 coding sequence in cases and none in over 7 , 000 controls . These deletions ranged in size from 54 kb to over 2 Mb ( Table 4 ) . DMRT1 is situated in a region of chromosome 9p that has been identified as a source of syndromic and non-syndromic forms of XY gonadal dysgenesis ( GD ) . The deletions of this region that are associated with syndromic forms of GD are usually 4–10 Mb in size , while isolated GD has been reported for deletions smaller than 1 Mb [40] , [51] , [52] . Despite frequent involvement of DMRT1 in these putative causal mutations , there is variability in both the phenotypic outcome affiliated with each deletion and the extent of DMRT1 coding sequence contained therein . At least two cases of GD have been linked to deletions near but not overlapping DMRT1 – one 700 kb mutation 30 kb distal to DMRT1 in a case of complete XY GD that was inherited from an apparently normal mother , and a second 260 kb de novo deletion about 250 kb distal to DMRT1 [39] , [40] . Both of these deletions overlapped the genes KANK1 and DOCK8 . On the other hand , two smaller deletions , one a 25 kb deletion of DMRT1 exons 1 and 2 , and one a 35 kb deletion of exons 3 and 4 , have been observed in patients with complete GD and bilateral ovotesticular disorder of sexual development , respectively [51] , [52] . Based on the clinical records of patients in our current study , there is no chance that our DMRT1 deletion carriers could represent misdiagnosis of a condition as severe as complete XY GD , which presents with the appearance of female genitalia . Indeed , two of our DMRT1 deletion carriers were subject to testicular biopsies . Our observations here suggest that hemizygous deletion of DMRT1 is a lesion that shows variable expressivity that may depend on the sequence of the undeleted DMRT1 allele , variation in other sequences on chromosome 9p , and the state of other factors in the pathways regulating testicular development and function . Strictly speaking , statements that hemizygous deletions of DMRT1 are “sufficient” to cause GD or spermatogenic failure need to be qualified at this point until we gain a better understanding of the effects of genetic background . For instance , in most studies of DMRT1 deletion , the undeleted DMRT1 allele is rarely sequenced . Is the mode of action dominant or recessive ? Deletions of the Y chromosome have long been appreciated as a cause of azoospermia , and we have now shown here that Y-linked duplications are also significant risk factors for spermatogenic failure . The precise definition of the duplication sensitive sequences awaits further investigation . Historically , Y duplications have been much less studied than Y deletions , as +/− STS PCR is the standard assay for assessing Y chromosome copy number variation in both the clinical and research setting . Quantitative PCR methods for measuring Y chromosome gene dosage have been described in the literature , and applied almost exclusively to studying the phenotypic effects of duplication of genes in the AZFc region [53] . Results of these investigations are conflicting , with studies of Europeans reporting no association between AZFc partial duplication and spermatogenic impairment [54] , while reproducible associations have been reported in east Asian cohorts [55] , [56] . Notably , we identified some duplications on the Y chromosome greater than 2 . 5 Mb in size , all spanning the AZFc locus ( Figure S6 ) , in 8/179 cases ( those typed on Affymetrix 6 . 0 ) , compared to 13/972 controls ( OR 3 . 45 [1 . 21–9 . 12] , p<0 . 01 ) . Rearrangements of this size on the autosomes confer staggering risk for other forms of disease; for example , by one recent estimate CNVs larger than 3 Mb have an OR of 47 . 7 for intellectual disability and/or developmental delay [46] . Our results suggest that Y chromosome structure may be more dosage sensitive than previously appreciated , and we speculate that some genes and non-coding sequences of the Y chromosome may be under stabilizing selection for copy number [57] . Three recent studies have used array-based approaches to characterize CNVs in men with azoospermia . Our finding of an X-linked CNV burden in men with spermatogenic failure has been replicated and described elsewhere [58] . In a second study , Tuttelmann et al . evaluated 89 severe oligozoospermic , 37 azoospermic , and 100 normozoospermic control men using Agilent 244K and 400K arrays and identified a number of CNVs potentially involved in male infertility [24] . Third , Stouffs et al . assayed nine azoospermic men and twenty control samples using the 244K array and followed-up CNVs of interest by q-PCR in up to 130 additional controls [25] . Using the criterion of at least 51% reciprocal overlap , we have identified a number of CNVs in the current study that overlap with case-specific CNVs in the Tuttelmann and Stouffs studies . The majority of these CNVs appear to be relatively common polymorphisms and not case-specific in our larger dataset; however several noteworthy CNVs overlap between studies and are absent , or are present at a very low frequency in controls . For example , Tuttelmann et al . identified a private duplication on Xq22 . 2 in an oligozoospermic man [24] , and we identified an overlapping duplication in an oligozoospermic man from the present study ( ChrX:103065826–103205985 , NCBI36 ) . These duplications alter the copy number of a small number of testis-specific or testis-expressed variants of histone 2B ( H2BFWT , H2BFXP , H2BFM ) . No CNVs in this region were identified in more than 1600 controls . Tuttelmann et al . also identified an azoospermic man with a deletion and another with a duplication on 8q24 . 3 , encompassing the genes PLEC1 and MIR661 [24] . We identified an oligozoospermic man with a duplication of the same region , affecting the same functional elements ( chr8:145064091–145118650 , NCBI36 ) . CNVs of this locus are very rare , with a frequency of about 0 . 005% in our controls and 0 . 0025% in controls used for a recent study of developmental delay [46] . It is important to note that new variants will frequently be discovered whenever a discovery technology such as array CGH is applied to a new sample set , and the observation that a variant is patient-specific is not in itself remarkable , especially when one is investigating very small sample sizes . Our observation of low deletion HI scores in cases raises a number of considerations for future studies of the genetics of spermatogenic impairment . We interpret low HI scores in cases as evidence against a widespread role for dominant , highly penetrant deletions in spermatogenic failure . It is possible that our case recruitment , which pre-screened for normal karyotype , may have removed all large HI score events; however our identification of two large HI deletions of WT1 and MAPK1 indicate otherwise ( Figure 3 ) . A second concern is that the data used to train the haploinsufficiency prediction algorithm is in part based on features of deletions known to cause dominant pediatric disease , and that an analogous approach trained on fertility phenotypes may lead to different conclusions . There are few examples of dominant loss-of-function mutations causing isolated infertility in humans and only 5 of the >200 mouse infertility mutants described in a previous review showed a phenotype in heterozygous form [45] , so fitting a model of a dominant infertility mutation may be challenging in the short term . Nonetheless , developing disease-specific pathogenicity scores for infertility phenotypes should be a priority . Despite the differences between the genetic signatures of spermatogenic impairment and severe developmental disease noted above , there are connections in their epidemiology . Recent results estimate a 9 . 9% rate of birth defects in children conceived by intracytoplasmic sperm injection ( ICSI ) , the technology typically employed for assisting cases of severe male factor infertility , which is an OR of 1 . 77 compared to unassisted reproduction [59] . Among several possible explanations for this finding , our data raise the possibility that mutations that compromise gonadal function may act pleiotropically to disrupt development in other tissues . A better understanding of the genetic basis of male infertility is urgently needed in order to improve risk assessment for couples considering assisted reproduction . Clinical genomics is a paradigm in need of robust applications , and our finding of a large CNV burden in cases suggest that some infertility mutations may have the high penetrance required for clinical utility . Indeed some mutation screens are already used clinically in the management of male infertility . Although the presence of azoospermia can be easily assessed using a standard laboratory test , many men with azoospermia will have sperm production within the testis and be candidates for testicular sperm retrieval . We have already identified that the specific AZF deletion ( a , b or b/c ) has a dramatic effect on the prognosis of sperm retrieval ( vs . AZFc-deleted males ) [60] . In the present study , we have identified deletion of DMRT1 coding sequence as a genetic event that appears highly predictive of spermatogenic failure . In depth characterization of carriers is now needed to understand how this mutation affects the prognosis of sperm retrieval . Similar whole genome tests may provide critical prognostic information that can help to characterize the chance of successful treatment for couples with non-obstructive azoospermia , avoiding expensive and needlessly invasive interventions , while potentially providing guidance for new therapeutic interventions .
All DNA samples used in this study were derived from peripheral blood lymphocytes collected from individuals giving IRB-approved informed consent . The following IRBs were involved: INSA Ethics Committee and Hospital Authority ( Portugal ) , University of Utah IRB , and Washington University in St . Louis IRB ( #201107177 ) . All samples of genomic DNA to be analysed in this study i ) belong to DNA banks that have been established throughout the years; ii ) are coded; and iii ) each individual has signed a declaration of informed consent before donating his genomic DNA for analysis , authorizing molecular studies to be performed with this material . All cases were deemed idiopathic following a standard clinical workup , which included screening for Y chromosome deletions . Controls from the Utah cohort were men with normal semen analysis , remaining controls were not phenotyped on semen quality . Full details of the source and diagnosis of samples in this study are available in Supplemental Methods . When using SNP arrays , CNV analysis is more sensitive to experimental noise than SNP genotyping , and we used different sample QC metrics to inform CNV and SNP stages of our project . As a result , we have slightly larger sample sizes for the HBD analyses than for the CNV analyses . The individuals studied here were sourced from diverse geographic locations ( Table 1 , Text S1 ) . All primary samples ( e . g . 323 cases and 1133 control samples subjected to whole-genome genetic analysis ) were of self-reported Caucasian ancestry , but it was necessary to take additional steps to control for population structure in all aspects of the analysis . First , genetic ancestry of each sample was assessed by principal components analysis and ethnicity outliers were removed ( Figure S2 , Figure S3 ) . Second , eigenvectors generated by this principal components analysis were used as covariates in both CNV association and inbreeding coefficient association analyses . For analyses focusing on the Y chromosome , we performed analyses conditioning on Y haplogroup to provide the most stringent possible correction for population structure with available data . Lastly , we conducted alternate association analyses with the Porto case cohort using a smaller , but more geographically proximal Spanish control cohort ( Figure S5 ) . Three array platforms were used for CNV discovery: Illumina 370K ( Utah ) , Illumina OmniExpress ( Washington University ) , and Affymetrix 6 . 0 ( Porto , Cornell , Nanjing ) . Full details of sample processing and array experiments are available in Supplemental Methods . Three CNV calling algorithms were used to generate CNV maps for each individual typed with Illumina technology: GADA , a sparse Bayesian learning approach [61]; PennCNV , a Hidden Markov Model ( HMM ) -based method originally designed for the Illumina platform [62]; and QuantiSNP 2 . 0 , another HMM-based method for Illumina [63] . CNVs called by 2 of 3 algorithms were retained for analysis . CNV calling for Affymetrix 6 . 0 was performed with Birdsuite [64] . Due to the complexity of calling CNVs on the sex chromosomes , for all array datasets we implemented a bespoke normalization and calling procedure that used only the GADA algorithm to call CNVs from the X and Y chromosomes . For full details of CNV calling see Supplemental Methods . Regions of homozygosity-by-descent ( HBD ) were identified using BEAGLE 3 . 0 [48] . SNPs with no-call rates >5% were removed prior to HBD analysis . As BEAGLE uses a model for background linkage disequilibrium that is fit from the data , cases and controls from each cohort were analyzed simultaneously and separately to assess cohort-specific biases in calling HBD . Prior to downstream analysis , we identified and removed a small number of reported HBD regions that corresponded to rare , large hemizygous deletions . Inbreeding coefficients for each individual were calculated from their HBD data using the formula: Due to differences in array content , CNV frequencies were determined on a per-platform basis . All CNV calls made on a given platform , in both cases and controls , were combined into CNV regions using a threshold of 50% reciprocal overlap to defined two events as the same ( [35] ) . We defined the CNV frequency as the proportion of all samples ( cases and controls ) containing that CNV . We constructed several statistical tests to measure differences between cases and controls . We used Mann-Whitney U tests to test for differences in the total amount of aneuploid sequence per genome . We used standard logistic regression to test for CNV load on chromosome compartments ( e . g . the autosomes , X chromosome ) and a small number of functional features ( genes , miRNA , etc ) . To control for population structure these models included the first 10 principal components from PCA analysis of the SNP genotype data from all cohorts ( Figure S2 ) . We used a permutation strategy for genomewide , locus-by-locus testing for association at all genes and in 500 kb non-overlapping genomic windows . The permutation strategy , implemented with the software package PLINK , calculates nominal and genomewide p-values by permuting case-control labels [65] . To present consistent summaries of CNV burden for the entire study ( all cohorts combined ) , we used linear mixed-effects logistic regression , treating cohort as a random factor and compared these to effect size estimates for each cohort separately using standard logistic regression ( Figure 1 ) . The mixed effects modeling framework controls for SNP platform as each case-control cohort was typed on a different platform; a similar use of mixed-effect modeling was recently described in a meta-analysis of schizophrenia SNP data [27] . Analogous tests were conducted on HBD segments from the original discovery cohort and the combined primary and replication datasets . We performed validation and replication analyses of DMRT1 deletions with and assay based on Taqman PCR . Copy number was assessed using a pre-designed assay #Hs06833797_cn within the DMRT1 gene against an RNase P reference ( assay # 4403326; both assays from Applied Biosystems , Carlsbad , CA , USA ) according to manufacturer's recommendations .
|
Infertility is a disease that prevents the transmission of DNA from one generation to the next , and consequently it has been difficult to study the genetics of infertility using classical human genetics methods . Now , new technologies for screening entire genomes for rare and patient-specific mutations are revolutionizing our understanding of reproductively lethal diseases . Here , we apply techniques for variation discovery to study a condition called azoospermia , the failure to produce sperm . Large deletions of the Y chromosome are the primary known genetic risk factor for azoospermia , and genetic testing for these deletions is part of the standard treatment for this condition . We have screened over 300 men with azoospermia for rare deletions and duplications , and find an enrichment of these mutations throughout the genome compared to unaffected men . Our results indicate that sperm production is affected by mutations beyond the Y chromosome and will motivate whole-genome analyses of larger numbers of men with impaired spermatogenesis . Our finding of an enrichment of rare deleterious mutations in men with poor sperm production also raises the possibility that the slightly increased rate of birth defects reported in children conceived by in vitro fertilization may have a genetic basis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"genome-wide",
"association",
"studies",
"aneuploidy",
"x-linked",
"chromosomal",
"disorders",
"genetics",
"biology",
"human",
"genetics",
"genetics",
"of",
"disease",
"chromosomal",
"deletions",
"and",
"duplications",
"genetics",
"and",
"genomics",
"y-linked"
] |
2013
|
Human Spermatogenic Failure Purges Deleterious Mutation Load from the Autosomes and Both Sex Chromosomes, including the Gene DMRT1
|
Adults with chronic Trypanosoma cruzi exhibit a poorly functional T cell compartment , characterized by monofunctional ( IFN-γ-only secreting ) parasite-specific T cells and increased levels of terminally differentiated T cells . It is possible that persistent infection and/or sustained exposure to parasites antigens may lead to a progressive loss of function of the immune T cells . To test this hypothesis , the quality and magnitude of T . cruzi-specific T cell responses were evaluated in T . cruzi-infected children and compared with long-term T . cruzi-infected adults with no evidence of heart failure . The phenotype of CD4+ T cells was also assessed in T . cruzi-infected children and uninfected controls . Simultaneous secretion of IFN-γ and IL-2 measured by ELISPOT assays in response to T . cruzi antigens was prevalent among T . cruzi-infected children . Flow cytometric analysis of co-expression profiles of CD4+ T cells with the ability to produce IFN-γ , TNF-α , or to express the co-stimulatory molecule CD154 in response to T . cruzi showed polyfunctional T cell responses in most T . cruzi-infected children . Monofunctional T cell responses and an absence of CD4+TNF-α+-secreting T cells were observed in T . cruzi-infected adults . A relatively high degree of activation and differentiation of CD4+ T cells was evident in T . cruzi-infected children . Our observations are compatible with our initial hypothesis that persistent T . cruzi infection promotes eventual exhaustion of immune system , which might contribute to disease progression in long-term infected subjects .
Chagas disease , a neglected tropical disease affecting approximately 10 million people from south of the United States to Mexico and Central and South America [1] , is caused by the protozoan parasite Trypanosoma cruzi . As a consequence of migration flows , the disease has been also become established in non-endemic countries [2] . T . cruzi frequently results in the development of cardiomyopathy , generally many years after the initial infection . Three factors are likely associated with the development of severe disease: parasite burden; the effectiveness of the host immune response in controlling parasites in specific tissues , and the effectiveness of the host immune response in limiting peripheral damage [3] , [4] . Chronic infections in general are associated with a progressive loss of pathogen-specific T cell function known as immune exhaustion [5] , [6] . We have previously shown that adults with chronic T . cruzi-infections exhibit a prevailing profile of parasite-specific Interferon ( IFN ) -γ only secreting T cells [7] , associated with long-term antigen persistence and exhausted T cells [5] . The frequencies of T . cruzi-specific T cells were also found to inversely correlate with the severity of chronic Chagas disease [7] , [8] . Although , T . cruzi-infected children are likely to have shorter-term infections than most adults , the overall CD8 T cell compartment in children in the early phase of chronic T . cruzi infection exhibits decreased levels of naïve T cells and increased levels of terminally differentiated antigen-experienced T cells [9] . Other studies have suggested that T . cruzi-infected children have a mixed T cell profile with the production of IFN-γ and IL-4 [10] , [11] . However , a comprehensive analysis of the ability of T . cruzi specific T cells to co-express multiple functions has not been performed . In order to examine the progression of immune exhaustion in chronic T . cruzi infection , we have assessed the quantitative and qualitative attributes of T . cruzi-specific T cell responses in early stages of Chagas disease in children in comparison to T . cruzi-infected adults . The degree of activation , differentiation and antigen exposure of total CD4+ T cells was also evaluated in T . cruzi-infected children . The primary finding is that T . cruzi-infected children have a higher frequency of polyfunctional and more robust T cell responses specific for T . cruzi compared to T . cruzi-infected adults , and a heightened state of immune activation of CD4+ T cells .
This protocol was approved by the Institutional Review Boards of the Hospital Interzonal General de Agudos Eva Perón , and the Centro Nacional de Genética , Buenos Aires , Argentina . Informed written consent was obtained from adult subjects and the parents of all children included in this study . Five to 16-year old children and 29 to 63 year-old adults were enrolled at the Instituto Nacional de Parasitología Dr . Mario Fatala Chaben and at the Hospital Interzonal General de Agudos Eva Perón , ( Buenos Aires , Argentina ) . T . cruzi infection was determined by indirect immunofluorescence , haemagglutination and ELISA assays [12] . Subjects positive on at least two of these tests were considered to be infected . All infected children were in the early chronic phase of T . cruzi infection . Age- and sex-matched children with negative serological findings were recruited as uninfected controls . T . cruzi-infected subjects were classified according to a modified version of the Kuschnir grading system [13] , [14] . Most subjects were in the indeterminate phase of the infection ( Group 0; positive serology and normal findings on electrocardiogaphy , chest radiography and echocardiography ) , while 3 subjects ( 2 adults and one child ) belong to the Group 1 group ( positive serology , abnormal findings on electrocardiography and normal chest radiography and echocardiography ) . All participants of this study had not received etiological treatment at the time of sampling and started drug therapy with benznidazole after sampling . Children and adults with any impaired health condition such as severe nephropathy , liver disease , severe neuropathy , severe anaemia or homeopathy were excluded from the study . Data on the number , sex , age , clinical and epidemiologic features of the subjects included in this study are summarized in Table 1 . Approximately 10 mL of blood were drawn by venipuncture into heparinized tubes ( Vacutainer; BD Biosciences ) . PBMC were isolated by density gradient centrifugation on Ficoll-Hypaque ( Amersham ) and were cryopreserved for later analysis . Additional 2 mL of blood were allowed to coagulate at 37°C and centrifuged at 1000 g for 15 min for sera separation . Protein lysate from T . cruzi amastigotes was obtained by freeze/thaw cycles followed by sonication as previously reported [8] . Tetanol Pur ( Novartis , Germany ) was used as source of tetanus toxoid . HLA-A01 , A02 , A03 , A24 and B44-supertype binding epitopes encoded by the trans-sialidase gene family of T . cruzi and peptides derived from Influenza ( Flu ) virus with high binding-affinity for the common class I HLA-supertypes A01 , A02 and A03 were synthesized at the University of Georgia Molecular Genetics Instrumentation Facility ( Athens , USA ) . The number of T . cruzi antigen-responsive IFN-γ- and IL-2-secreting T cells was determined by ex vivo ELISPOT using commercial kits ( BD Biosciences ) , as described elsewhere [7] , [8] , [15] . Cryopreserved PBMC were seeded in triplicate wells , at a concentration of 4×105 cells/well , and stimulated with T . cruzi lysate ( 10 µg/mL ) or with peptide pools from the trans-sialidase protein family ( 7–13 peptides per pool , 5 µg/ml/peptide ) . For controls , PBMCs were incubated with either 20 ng/mL phorbol 12-myristate 13-acetate ( PMA , Sigma ) plus 500 ng/mL ionomycin ( Sigma ) , Flu virus ( 3 peptides per pool , 5 µg/ml/peptide ) , tetanus toxoid ( 4 I . U . /ml ) or media alone ( unstimulated control ) for 16–20 hr at 37°C and 5% CO2 . Spot forming cells ( SFC ) were automatically enumerated using ImmunoSpot analyzer ( CTL ) . The mean number of spots in triplicate wells was obtained for each condition , and the number of specific IFN-γ and IL-2-secreting T cells was calculated by subtracting the value of the wells containing media alone from the antigen-stimulated spot count . Responses were considered significant if a minimum of 10 spots/4×105 PBMC were present per well , and additionally , this number was at least twice the value of wells with media alone [8] . For polyfunctional analysis , 200 µL heparinized blood , half diluted in RPMI , were incubated with T . cruzi lysate in the presence of anti-CD28 and anti-CD49d antibodies ( 1 µg/ml; BD Pharmingen ) , for 16–20 h at 37°C . Ten µg/ml brefeldin A ( Sigma ) were added for the last 5 h of incubation , as previously described [16] . Blood incubated with no antigen served as a negative control ( unstimulated control ) , while blood incubated with Staphylococcal enterotoxin B ( 1 µg/ml; Sigma-Aldrich ) served as a positive control . Twenty mM EDTA was added for 15 min . Cells were then stained with anti-human CD4-peridinin chlorophyll protein ( PerCP ) ensued by red cell lysis and white cell fixation in FACS Lysing Solution ( Pharmingen ) . This was followed by fixation and permeabilization with Cytofix/Cytoperm solution ( Pharmingen ) according to manufacturer's instructions . Cells were stained with a combination of monoclonal antibodies specific for IFN-γ [phycoerythrin ( PE ) ] , tumor necrosis factor ( TNF ) -α [allophycocyanin ( APC ) ] and CD154 ( CD40L ) [fluorescein isothiocyanate ( FITC ) ] , all from BD Bioscience . Since CD154 is only transiently expressed on the cell surface [17] , intracellular expression of CD154 was measured , as previously described [18] . In order to confirm that cytokine/co-stimulation expression was derived from T cells , anti-human CD3-FITC or CD3-PerCP was added in polyfunctional staining assays in combination with CD4 , IFN-γ and TNF-α or CD4 , IFN-γ and CD154 , respectively . Typically , 500 , 000 lymphocytes were acquired on a FACScalibur ( Becton Dickinson Immunocytometry Systems ) and analyzed using FlowJo software ( TreeStar , Inc . ) Lymphocytes were identified based on their scatter patterns and CD4 expression for the combination of IFN-γ , TNF-α and CD154; and based on scatter patterns as well as CD3 and CD4 expression for the combination of IFN-γ and TNF-α or IFN-γ and CD154 . Boolean combination gating was then performed to calculate the frequencies of expression profiles corresponding to the seven different combinations of functions by using the FlowJo software . After subtracting the background values , the proportions of the different subsets were expressed as percentages of total cytokine or CD154-positive cells . Responses to the T . cruzi lysate were considered positive , for any particular subset , if the frequency of cytokine/CD154-positive T cells was threefold higher than the frequency in medium alone and above 0 . 07% of total CD4+ T cells , since the limit of detection was set at 0 . 01% . For phenotypic analysis of total CD4+ T cells , 50 µL of whole blood were incubated with different combinations of CD4 ( PerCP ) , CD45RA ( FITC or APC ) , CD27 ( PE ) , CD28 ( PE ) , CD127 ( PE ) , HLA-DR ( FITC ) and KLRG1 ( APC ) monoclonal antibodies ( BD Biosciences ) , followed by red cell lysis and white cell fixation in FACS Lysing Solution ( Pharmingen ) . Typically , 500 , 000 lymphocytes were acquired on the FACScalibur ( Becton Dickinson Immunocytometry Systems ) and analyzed using FlowJo software ( TreeStar , Inc . ) Demographic and clinical characteristics of T . cruzi-infected subjects included in this study were summarized using range and median and compared with healthy participants using Kruskal-Wallis test with Dunn correction and Fisher's exact tests for numerical and categorical variables , respectively . Comparisons of the frequencies of responders in the different categories of ELISPOT responses among T . cruzi-infected children and between T . cruzi-infected children and the uninfected group were evaluated by use of the chi 2 test and Fisher's exact test . To compare the number of spots , as well as the magnitude of CD4+ T cells responses among triple , double or single functional profiles , the Kruskal-Wallis nonparametric analysis with Dunn correction was used . The percentages of responding CD4+ T cells for each cytokine within and among the children and adult groups were compared by the Kruskal-Wallis nonparametric analysis with Dunn correction and Mann-Whitney U test , respectively . T cell phenotypes in T . cruzi-infected children and uninfected controls were compared applying the Mann-Whitney U-test . Differences were considered to be statistically significant at P<0 . 05 .
The main characteristics of the entire study population are summarized in Table 1 . T . cruzi-infected and uninfected children enrolled in the study were all born from mothers with positive serology for T . cruzi infection and , at the time of the study all children were living in Buenos Aires , where T . cruzi infection is not endemic . T . cruzi-infected and uninfected groups of children included individuals born in areas endemic for T . cruzi infection and individuals born in non-endemic areas . Seropositive children who had not lived in endemic areas are presumed to have been infected congenitally . Only one child out of the fifty T . cruzi-infected children showed a right bundle branch block that is among the electrocardiographic alterations related to Chagas disease . Abnormal findings in the electrocardiography not related to Chagas disease , including sinusal tachycardia , incomplete right bundle branch block and premature atrial contraction [19] , were more frequent among T . cruzi-infected children born in areas endemic for T . cruzi infection than those born in non-endemic areas ( Table 1 ) . The presence of these electrocardiographic alterations were not correlated with the age or gender of children . All T . cruzi-infected adults were born in areas endemic for T . cruzi infection but have lived in Buenos Aires for more than 15 years , in average . Two adult patients belonged to the G1 clinical stage of the Kuschnir classification [13] , showing left anterior fascicular block while the remaining 8 subjects belong to the G0 group . We have previously reported an inverse correlation between the frequency of IFN-γ-producing T cells responsive to T . cruzi antigens and disease severity in chronically infected adults [7] , [8] . In this study , the magnitude of T . cruzi-specific T cell responses was assessed in 17 T . cruzi-infected children by measuring the total frequency of IFN-γ- and IL-2-producing cells by ELISPOT assays after stimulation of PBMC with T . cruzi antigens . Since we had shown that in chronically T . cruzi-infected humans , the frequency of T cells specific for class-I-restricted T . cruzi epitopes ( CD8+ epitopes ) [7] , [8] or T . cruzi–derived recombinant proteins [20] is low to be consistently detected , an amastigote lysate preparation was the primary antigen stimulus for ELISPOT assays . More than 75% of infected children had detectable cytokine responses to T . cruzi-antigens , with secretion of IFN-γ and IL-2 in nearly 60% of T . cruzi-infected children and , approximately 18% with IFN-γ-only responses . IL-2-only responses were not observed in T . cruzi-infected children ( Table 2 ) . None of the seronegative children born from mothers with positive serology for T . cruzi infection had positive ELISPOT responses ( Table 2 ) . The number of IFN-γ spots in subjects with concomitant IFN-γ and IL-2 positive ELISPOT responses was higher than the number of IFN-γ spots in IFN-γ-only responders ( Figure 1A ) . Likewise , the former group showed higher spot counts for IFN-γ than IL-2 PBMC available from five T . cruzi-infected children were also stimulated with peptides encoded by the trans-sialidase gene family that bind alleles representative of the 6 most common class I HLA-supertypes , previously shown as targets of CD8+ T cell responses in chronically T . cruzi-infected adults [7] . Strong IFN-γ ELISPOT responses specific for five out of the six supertype-binding trans-sialidase pooled peptides assessed were observed in one out of the five children ( Figure 1B ) . In this same child , strong IL-2 responses to HLA-A3 and HLA-A24-binding peptides were also detected ( Figure 1B ) . In opposition to T . cruzi-specific T cell responses , the magnitude of IFN-γ ELISPOT responses to irrelevant antigens , including class-I-restricted Flu peptides and tetanus toxoid , did not differ among uninfected children , uninfected adults and T . cruzi-infected adults ( Figure 2 ) , supporting the impairment of T cell responses is restricted to T . cruzi . To further determine the quality of T . cruzi specific T cell responses earlier in infection , cytokine/co-stimulation co-expression profiles were assessed by intracellular staining following stimulation of whole blood with T . cruzi lysate . Since the T . cruzi lysate has proved to mainly induce CD4+ T cell responses and to a much lesser extent CD8+ T cell responses in chronic T . cruzi-infections [7] , [8] , we focused on CD4+ T cell responses for flow cytometry assays . The ability of CD4+ T cells to produce IFN-γ , TNF-α and CD154 ( CD40L ) - a marker of costimulatory potential - was evaluated in 19 T . cruzi-infected children . For comparison , this functional profile was also assessed in 10 T . cruzi-infected adults without evidence of cardiac involvement or with mild cardiac disease . Examining IFN-γ , TNF-α and CD154+ individually , CD154 was the most highly expressed in T . cruzi-infected children and there was a trend for higher proportions of cells expressing one or more of these three markers in the in the T . cruzi-infected children relative to the adults ( Figure 3 and Figure S1 ) . IFN-γ-producing CD4+ T cells was prevalent among T . cruzi-infected adults ( Figure 3 ) . A Boolean gating analysis was then performed to categorize cytokine/CD154 positive cells into seven different subsets consisting of triple , double or single cytokine/CD154-expressing populations . T . cruzi antigen responsive CD4+ T cells in children with positive serology for T . cruzi infection had representatives of all seven possible subsets ( range 1–5 subsets per subject ) ( Figure 4A and 4C ) . CD4+ T cells responding to the lysate exert at least 2 functions in >90% of T . cruzi-infected children ( Figure 4D ) . Polyfunctional cytokine-producing cells that simultaneously secrete IFN-γ and TNF-α dominated the CD4+ T cell responses in T . cruzi-infected children , followed by monofunctional and double expressing IFN-γ+CD154+ CD4+ T cells . TNF-α+CD154+ and triple expressing IFN-γ+TNF-α+CD154+ cells represented , in average , less than 20% of the total CD4+ T cell response to T . cruzi-antigens ( Figure 4A and 4C ) . In contrast to children , CD4+ T cell responses in infected adults had representatives of only 3 out of the 7 possible subsets ( Figure 4B and 4C ) , with a higher proportion of subjects displaying CD4+ T cells with single function than double function and none with all three functions ( Figure 4D ) . In adult patients , monofunctional responses were dominated by IFN-γ-producing T cells followed by CD154+ T cells , whereas IFN-γ+CD154+ cells were predominant among the subsets with double function ( Figure 4B ) . Monofunctional CD154+ followed by monofunctional IFN+ T cells and double expressing IFN-γ+CD154+ T cells showed the highest magnitude of CD4+ T cell responses to T . cruzi antigens in T . cruzi-infected children ( Figure S2A ) . Likewise , monofunctional T cells are also the main contributors to the total magnitude of CD4+ T cell responses to the lysate in T . cruzi-infected adults , but contrasting with children , T cell responses were enriched in single IFN-γ+ rather than single CD154+ T cells ( Figure S2B ) . We have previously reported that T . cruzi-infected children exhibit increased levels of terminally differentiated CD8+ T cells in the overall T cell compartment [9] . In this study , the phenotype of total CD4+ T cells was evaluated for the expression of CD45RA , the co-stimulation molecules CD27 and CD28 , the lymph node homing receptors CCR7 and CD62L , the IL-7 receptor CD127 and the co-inhibitory receptor killer-cell lectin like receptor G1 ( KLRG-1 ) , a marker of the number of TCR-triggering events [21] , [22] . Highly differentiated memory T cells ( CD45RA−CD27−CD28− and CD45RA−CCR7−CD62L− ) and terminally differentiated effector T cells ( CD45RA+CCR−CD62L− ) are increased in T . cruzi-infected children compared with uninfected controls ( Table 3 ) . The increased expression of HLA-DR indicates a high degree of activation of CD4+ T cells , while the increased expression of KLRG-1 along with a loss of the IL-7 receptor , suggests sustained antigen exposure of T cells in T . cruzi-infected children .
T cell deficiencies resulting from exhaustion include a hierarchical loss of effector functions along with the expression of inhibitory receptors , and failure to exhibit antigen-independent homeostatic proliferation [5] , [23] , [24] . Although , this phenomenon was initially described for chronic viral infections , recent studies have shown that this process can also occur in protozoan diseases [6] , [25] . We have previously provided several pieces of evidence of immune exhaustion in adults with chronic T . cruzi-infection [8] , [26]–[29] . In this study , we report for the first time that T . cruzi-infected children in early stages of T . cruzi infection maintained polyfunctional CD4+ T cells responsive to T . cruzi antigens in their circulation . More than ninety percent of T . cruzi-infected children displayed responsive CD4+ T cells with double or triple functions . These findings are in sharp contrast to the profile observed in a group of adults with chronic T . cruzi-infections , in which most subjects showed monofunctional T . cruzi-specific responses enriched in single IFN-γ+ T cells with absence of TNF-α+-producing T cells . Other authors had also reported that children in the early phase of chronic T . cruzi infection exhibit IFN-γ+- and TNF-α+-secreting CD4+ T cells in response to T . cruzi antigens [10] , [11] . The analysis of IFN-γ ELISPOT responses to the lysate and to trans-sialidase-derived class I peptides in T . cruzi-infected children also revealed that not only were the proportion with polyfunctional responses higher relative to infected adults , but that the magnitude of IFN-γ responses in these positive children was higher by 5-fold than previously reported in chronically T . cruzi-infected adults ( i . e . median number of IFN-γ spots/4×105 PBMC in children = 220 vs . 52 spots evaluated in 150 subjects in the G0 group ) , [Laucella] , [ unpublished data , 7 , 8] . Of note , the one single child responsive to trans-sialidase-derived CD8+ target peptides showed a stronger and broader response compared with our previous findings in adults ( i . e . median number of IFN-γ spots/4×105 in this child = 87 spots , with five out of the six supertype-binding trans-sialidase pooled peptides recognized in average vs . 19 spots in 25 T . cruzi-infected adults , with two out of six pooled peptides recognized in average ) [7] . The differences are even more dramatic when comparing T . cruzi-infected children with adults exhibiting severe cardiomyopathy who mostly showed negative T cell responses against T . cruzi antigens [7] , [8] . Recent studies have also shown that individuals with severe cutaneous leishmaniasis have reduced CD8+ T cell numbers with the ability to produce IFN-γ and IL-2 [30] . Multifunctional T cells are optimized for effector function , exhibiting a higher secretion of IFN-γ on a per-cell basis , more efficient killing , and IL-2-mediated expansion of T cells in an autocrine or paracrine manner [31] . In agreement with these findings , we observed a higher level of IFN-γ production in T . cruzi-infected children displaying both IFN+ and IL-2+ ELISPOT responses compared with those with single IFN-γ-producing T cells . It is likely that the maintenance of polyfunctional T cell responses provides more efficient control of T . cruzi infection and limits peripheral tissue damage . Indeed , most T . cruzi-infected children have not developed cardiac disease and previous studies from our group have shown higher frequencies of IFN-γ-producing T cells and higher magnitude of these responses in subjects with less severe forms of the disease compared with patients with severe cardiomypathy [7] , [8] , [29] . Of note , seropositive children who were born and lived for some years in areas endemic for T . cruzi infection showed higher electrocardiographic abnormalities not related to Chagas disease compared to children born in non-endemic areas . This is likely due to co-morbidities and poorer socioeconomic conditions in rural areas . In this respect , we have shown that socioeconomic conditions had a significant influence on the progression of chronic Chagas disease which was independent of antiparasitic treatment and clinic characteristics [32] . Nevertheless , the magnitude and quality of T . cruzi-specific T cell responses were not different between seropositive children born in endemic areas and those born in non-endemic areas , who are presumed to have been infected congenitally . In T . cruzi infection , parasite persistence for decades , might drive a process of immune exhaustion . A broader range ( relative to adults ) of T . cruzi-specific immune phenotypes are present in T . cruzi-infected children – from “adult-like” monofunctional CD4+ T cell subsets to T cells exhibiting up to 3 functional properties . Thus , some T . cruzi-infected children display a T cell profile similar to that of adult patients . This is not unexpected since some children will likely have infections of >10 years . The increased frequencies of total effector/effector memory and activated CD4+ T cells in T . cruzi-infected children observed herein suggest a heightened state of immune activation as previously suggested in other studies [9] , [33] , [34] In contrast to the unaltered levels of naïve CD4+ T cells shown herein , previous studies by our group in a similar patient population of children demonstrated a significant decrease in the levels of naïve CD8+ T cells compared to uninfected children [9] . Likewise , Vitelli-Avelar reported decreased levels of CD62L+CD4+ T cells but lower frequencies of CD4+HLA-DR+ T cells along with unaltered levels of CD4+CD28+ T cells [35] in children in the indeterminate phase of the infection . Adults with chronic T . cruzi infection also show increased levels of activated , late differentiated memory and effector T cells , while naïve T cells are diminished as disease becomes more severe [26] , [27] , [29] , [36] . A crucial question remains whether the heterogeneity in the functional and phenotypic profile is a predictor of different disease outcome in these patients . Children have not only shorter term infections but are also younger . Aging of the immune system in elderly subjects ( generally older than 65 years of age ) , known as immunosenescence [37] , is mainly associated to poor responsiveness to new pathogens and reduced efficacy of vaccination-induced protection against infection , while established memory immune responses to previously encountered pathogens are much less affected unless it is additionally stressed by chronic infections or autoimmune diseases [38] , [39] . Several studies have shown accelerated immune aging in HIV-infected subjects and subjects chronically infected with cytomegalovirus [40] . In this study , chronically T . cruzi-infected subjects are younger than 60 years , in average , and there were not differences in the magnitude of IFN-γ ELISPOT to irrelevant antigens like toxoid tetanus and Flu among uninfected children , uninfected adults T . cruzi-infected adults , . Of note , in previous studies , we did not found any association between the magnitude of T cell responses specific for T . cruzi and age in a cohort of chronically T . cruzi-infected subjects in an age range of between 22 and 70 years [7] , [8] . Furthermore , CD4+ T cells responsive to T . cruzi-antigens in T . cruzi-infected adults display high expression of the inhibitory receptor CTLA-4 which is a feature of exhausted T cells [41] , while few T cells responsive to tetanus and diphtheria toxoids expressed CTLA-4 [29] , supporting that impairment in T cell responses in T . cruzi-infected adults was confined to those specific for T . cruzi . The phenotype of the overall T cell compartment with an expansion of end-differentiated effector T cells and loss of naïve T cells in T . cruzi-infected children and adults compared with age-matched controls suggest that chronic T . cruzi infection might accelerate immune aging along with the induction of T cell exhaustion . The quality of the immune response might also be a factor in the higher cure rates following drug treatment observed in children relative to adults [42] . It had been proposed that IFN-γ production might promote the effectiveness of chemotherapy in patients with chronic T . cruzi infections [12] , [43] . All the children included in this study were treated with benznidazole , and are currently under follow-up . It will be of interest to address whether the response to therapy is associated with the qualitative and quantitative characteristics of the anti-T . cruzi T cell response prior to treatment .
|
Chagas disease is a neglected tropical disease affecting approximately 10 million people in the world . As a consequence of migration flows , the disease has also become established in non-endemic countries . Previous studies have demonstrated that Trypanosoma cruzi-specific T cells inversely correlates with the severity of cardiac disease in the chronic phase of the infection , suggesting that the immune system becomes exhausted overtime . To test this hypothesis , the quality and magnitude of T . cruzi-specific T cell responses were measured in T . cruzi-infected children – who are presumed to have shorter-term infections - and compared with long-term T . cruzi-infected adults . The activation status of total T cells in T . cruzi-infected children was also evaluated . T . cruzi-infected children exhibit a more robust , and more highly functional parasite specific T cell responses compared to T . cruzi-infected adults . In spite of a more functional immune profile , T . cruzi-infected children have a heightened state of immune activation . These observations are compatible with the initial hypothesis that T cell responses specific for T . cruzi become exhausted overtime . The impairment in T cell responses might contribute to disease progression in long-term infected subjects .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[] |
2013
|
Polyfunctional T Cell Responses in Children in Early Stages of Chronic Trypanosoma cruzi Infection Contrast with Monofunctional Responses of Long-term Infected Adults
|
Modern optical imaging experiments not only measure single-cell and single-molecule dynamics with high precision , but they can also perturb the cellular environment in myriad controlled and novel settings . Techniques , such as single-molecule fluorescence in-situ hybridization , microfluidics , and optogenetics , have opened the door to a large number of potential experiments , which begs the question of how to choose the best possible experiment . The Fisher information matrix ( FIM ) estimates how well potential experiments will constrain model parameters and can be used to design optimal experiments . Here , we introduce the finite state projection ( FSP ) based FIM , which uses the formalism of the chemical master equation to derive and compute the FIM . The FSP-FIM makes no assumptions about the distribution shapes of single-cell data , and it does not require precise measurements of higher order moments of such distributions . We validate the FSP-FIM against well-known Fisher information results for the simple case of constitutive gene expression . We then use numerical simulations to demonstrate the use of the FSP-FIM to optimize the timing of single-cell experiments with more complex , non-Gaussian fluctuations . We validate optimal simulated experiments determined using the FSP-FIM with Monte-Carlo approaches and contrast these to experiment designs chosen by traditional analyses that assume Gaussian fluctuations or use the central limit theorem . By systematically designing experiments to use all of the measurable fluctuations , our method enables a key step to improve co-design of experiments and quantitative models .
Stochastic gene expression can be modeled as a discrete state , continuous time Markov process , where different states x i = [ η 1 , η 2 , … , η N s ] i T ∈ X ⊂ Z ≥ 0 N s represent the Ns species of interest . In a biological context , the species η often correspond to gene configurations , RNA or protein abundances . Transitions to state xi + ψν from xi occur with probabilities wν ( xi , t ) dt in an infinitesimal time step of length dt , where wν and ψν are the propensity function and the stoichiometric vector corresponding to reaction ν ∈ {1 , 2 , … , Nr} . Using the propensity functions and stoichiometry vectors , one can describe the evolution of probability mass for each xi using the chemical master equation ( CME , [25 , 28] ) given by: d d t p ( x i ; t ) = ∑ ν = 1 N r [ w ν ( x i - ψ ν , t ) p ( x i - ψ ν ; t ) - w ν ( x i , t ) p ( x i ; t ) ] . ( 1 ) By enumerating all possible xi , one can define the probability mass vector as p = [p ( x1;t ) , p ( x2;t ) , …]T and reformulate the CME in matrix form as d d t p ( t ) = A p ( t ) [27] . Many systems described by the CME are not closed , i . e . the vector p has infinite dimension . In such cases , most states are extremely rare , and the sum of their corresponding probabilities is negligible . Thus , a natural approximation for the CME is to separate it into two exhaustive and disjoint sets , XJ and XJ′ , with XJ being a finite set and XJ′ being a set of low probability states . Defining pJ ( t ) ≡ p ( XJ; t ) , the CME can be reordered and written as: d d t ( p J ( t ) p J ′ ( t ) ) = ( A J J A J J ′ A J ′ J A J ′ J ′ ) ( p J ( t ) p J ′ ( t ) ) . ( 2 ) The finite state projection ( FSP ) approach [27] , obtains an approximation of pJ ( t ) for finite times by replacing the set of states XJ′ ( t ) with an absorbing sink state whose probability mass is g ( t ) , d d t ( p F S P ( t ) g ( t ) ) = ( A J J 0 - 1 T A J J 0 ) ( p F S P ( t ) g ( t ) ) . ( 3 ) The FSP provides the exact total error of this approximation for all states in XJ and XJ′ as: | ( p J ( t ) p J ′ ( t ) ) - ( p F S P ( t ) 0 ) | 1 = g ( t ) , ( 4 ) where the | . |1 denotes the absolute sum of the vector [24 , 27] . The FSP solution is also guaranteed to be a lower bound on the true solution [24 , 27] , ( p F S P ( t ) 0 ) ≤ ( p J ( t ) p J ′ ( t ) ) for all t > 0 . ( 5 ) For simplicity , we will hereafter refer to the approximated states pFSP ( t ) as p ( t ) and the corresponding matrix AJJ as A . Next , we derive the likelihood function for FSP models and single-cell data . A common task in single-cell analyses is to analyze snapshot measurements of independent cell populations , such as those collected using single-molecule fluorescent in-situ hybridization ( smFISH ) [22 , 23] . For such measurements , cells are fixed in the process of quantifying their RNA , and individual cells cannot be tracked over time . However , snapshots can be collected at different points in time to quantify a population’s response to changing conditions [2 , 29 , 30] . For such experiments , we assume that measurements at all time points {tk} are independent . The measured RNA counts for Ns different labeled species for each of Nc individual cells at time t can be collected into the data matrix D t ≡ [ d 1 , d 2 , … , d N c ] t ∈ Z ≥ 0 N s × N c . We define L ( D;θ ) as the likelihood that all measured data D = { D 1 , … , D N t } come from a model parameterized by θ = [θ1 , θ2 , … , θk] . For FSP models , the likelihood and its logarithm for Nc measured cells can be written directly as: L ( D ; θ ) = ∏ k = 1 N t ∏ i = 1 N c ( k ) p ( x i = d i ; t k , θ ) , ( 6 ) logL ( D ; θ ) = ∑ k = 1 N t ∑ i = 1 N c ( k ) log ( p ( x i = d i ; t k , θ ) ) . ( 7 ) A common task in systems biology is to estimate parameters θ ^ that maximize the likelihood that data could have come from a given model , and this form of the likelihood function has been used multiple times to estimate parameters from single-cell data [2 , 6 , 21 , 24 , 31 , 32] . In addition to estimating parameters from data , the likelihood function can also be used to estimate the sensitivity of parameter estimates to sampling errors in the experimental measurements , which can in turn be used to design better experiments . In the following sections , we will use this fact to derive the FIM for FSP models . The FIM , which describes the amount of information that can be expected by performing a particular experiment with Nc cells , is defined as I ( θ ) = N c E { ( ∇ θ log p ( X ; θ ) ) T ( ∇ θ log p ( X ; θ ) ) } , ( 8 ) where the expectation is taken over p ( X; θ ) , corresponding to the density from which future ( or hypothetical ) data could be sampled . For FSP models , this density is the discrete distribution found by solving Eq 3 . Eq 8 is positive semi-definite and is additive for collections of independent observations [10] . The inverse of the FIM is known as the Cramèr-Rao bound ( CRB ) , which provides a useful lower bound on the variance for any unbiased estimator of model parameters [11] . The notion of information stems from the fact that new experiments should increase the FIM , corresponding to additional knowledge about θ and a tighter CRB . More specifically , the well-known asymptotic normality of the maximum likelihood estimator ( MLE ) states that as the number of measurements Nc increases , the MLE estimates will converge in distribution to a multivariate normal probability density with a variance given by the CRB , N c ( θ ^ - θ * ) → d i s t N ( 0 , I ( θ * ) - 1 ) , ( 9 ) where θ ^ is the θ that maximizes Eq 6 and θ* are the “true” model parameters that produced the observed data [10 , 11] . Designing experiments to maximize a given metric of the FIM can be expected to provide a more accurate estimate of θ , where different definitions of ‘accuracy’ ( i . e . , different vector norms for parameter errors ) can be implemented through the choice of different FIM metrics . To derive the FIM requires one must take the partial derivative of the log-likelihood ( Eq 7 ) with respect to the parameters θ , ∇ θ logp ( X ; θ ) = ( 1 p 0 ∂ p 0 ∂ θ 1 1 p 0 ∂ p 0 ∂ θ 2 ⋯ 1 p 0 ∂ p 0 ∂ θ N p 1 p 1 ∂ p 1 ∂ θ 1 1 p 1 ∂ p 1 ∂ θ 2 ⋯ 1 p 1 ∂ p 1 ∂ θ N p ⋮ ⋮ ⋯ ⋮ 1 p N ∂ p N ∂ θ 1 1 p N ∂ p N ∂ θ 2 ⋯ 1 p N ∂ p N ∂ θ N p ) . ( 10 ) The expression ∇θ p ( X; θ ) is the sensitivity matrix , S , which has dimensions N × Nθ , where N is the dimension of the CME or its FSP projection . As described in the Materials and Methods , we derive an equation similar to that presented in [33] to define the time evolution of the sensitivity for each state’s probability density , p ( xl; θ ) , to each parameter θj . However , unlike previous analyses that rely on stochastic simulations and finite difference approaches , the FSP enables direct approximation of the sensitivities . Using the sensitivity matrix , the entries of the FIM can be computed as: I ( θ ) i j = N c E { ( 1 p ( x l ; θ ) ) 2 S l i S l j } . ( 11 ) Taking the expectation over all l on ( 1 , N ) yields the elements of the FIM: I ( θ ) i j = N c ∑ l = 1 N ( 1 p ( x l ; θ ) ) 2 S l i S l j p ( x l ; θ ) , = N c ∑ l = 1 N 1 p ( x l ; θ ) S l i S l j , ( 12 ) which quantifies Fisher information for the model evaluated at a single time point . For smFISH data , each time point is independent . If Nc ( tk ) cells are measured at each kth time point , the FIM is summed , and the total information is computed as: I ( θ ) i j = ∑ k = 1 N t N c ( t k ) ∑ l = 1 N 1 p ( x l ; t k , θ ) S l i ( t k ) S l j ( t k ) . ( 13 ) The Fisher information can be found using Eq 13 for any model for which the FSP ( Eq 3 ) can be solved . This formulation explicitly quantifies how the number of cells and number of time points impact the information , and is easily extended to include other experiment design aspects such as the interval of successive measurements or changes in applied inputs , as we will demonstrate in the following sections . Because one is often interested in the relative sensitivity of parameters rather than the absolute sensitivity , a logarithmic parameterization of the FIM can easily be obtained from Eq 13 by multiplying by the corresponding entries of θ ( see supplemental information for full details ) , I ( log θ ) i j = θ i θ j I ( θ ) i j . ( 14 ) In the following sections , we will verify the FIM using several common models of gene expression , and demonstrate experiment designs using these approaches .
To demonstrate and validate the FSP-FIM method , we begin with a simple birth and death model for constitutive gene expression as shown in Fig 1 . This model , which has been fit to capture the variability for many housekeeping genes [1 , 20] , consists of two reactions , corresponding to the constant transcription and first order decay of RNA , R 1 : g e n e → k r g e n e + R N A R 2 : R N A → γ ∅ . The production and degradation parameters are defined as θ = [kr , γ] . Given an initial condition of zero RNA for this process , the population of RNA at any later time is a random integer sampled from a Poisson distribution , p ( x ; λ ) = λ x e - λ x ! , ( 15 ) where λ is the time varying average population size , λ ( t , k r , γ ) = k r γ [ 1 - exp ( - γ t ) ] . ( 16 ) We have chosen the constitutive gene expression model to verify the FSP-FIM because the exact solution for the Fisher information for Poisson fluctuations can be derived in terms of λ as [10]: I Poisson ( λ ) = 1 λ . ( 17 ) For convenience , the derivation of Eq 17 is included in the supplementary text . Fig 1 shows the exact value of Fisher information ( orange ) versus the mean expression level for the two parameters kr and γ . Fig 1 also shows that the FSP-FIM ( blue ) matches the exact solution for the information on both parameters at all expression levels , which verifies the FSP-FIM for this known analytical form . Having demonstrated that the FSP-FIM matches to the exact solution , it is instructive to compare how well the previous LNA-FIM and SM-FIM estimates match to the exact FIM computation . For the Poisson distribution , the mean and variance are both equal to λ . Using this fact , the FIM can be approximated using the LNA-FIM for normal distributions ( see Eq 37 in the Materials and methods ) . This expression , which is derived in the supplementary text , reduces to I N ( λ , λ ) = 1 λ + 1 2 λ 2 , ( 18 ) when both the mean and variance are λ . As λ becomes large , the Poisson distribution becomes well approximated by a normal distribution [11] . Eqs 17 and 18 show that for this limit of large λ , the first term in Eq 18 dominates , and I N reduces to I Poisson , yielding nearly equivalent values for the expected information . However at low mean expression λ ≤ 1 , the strictly positive Poisson and the symmetric Gaussian distributions are less similar , and the Gaussian approximation predicts more information than is actually possible given the exact Poisson distribution . These trends are shown in Fig 1 , where the LNA-FIM approach only matches to the exact solution at high expression levels ( compare orange and purple lines ) . For this example , the sample-moments based FIM ( SM-FIM ) is exact and matches to the analytical and FSP-FIM solutions at all expression levels [9] . Next , we consider a slightly more complicated model of bursting gene expression , in which a single gene undergoes stochastic transitions between active and inactive states with rates kon and koff . This switching model , which is depicted in Fig 2 ( a ) , has been studied in detail [20 , 34–40] , and it has been used to capture single-cell smFISH measurements in mammalian cells [30 , 37] , yeast cells [2 , 36] , and bacterial cells [29] . When active , the gene transcribes RNA with constant rate kr and these RNA degrade in a first order reaction with rate γ . The four reactions of the system are: R 1 : g off → k on g on ( 19 ) R 2 : g on → k off g off ( 20 ) R 3 : g on → k r g on + R N A ( 21 ) R 4 : R N A → γ ∅ . ( 22 ) For the examples below , we use the baseline parameters given by: kon = 0 . 05α min−1 , koff = 0 . 15α min−1 , kr = 5 . 0 min−1 , and γ = 0 . 05 min−1 . In particular , the mRNA degradation rate , which sets the overall time-scale , was chosen to be representative of the average decay times ( approximately 20 minutes ) for mRNA in yeast [41] . For the bursting gene expression model , rescaling the transition rates kon and koff by a common factor does not affect the mean expression level , because the fraction of time spent in the active state remains unchanged . This fraction can be written f on ≡ α k on α k on + α k off = k on k on + k off , ( 23 ) and is the same for any α > 0 . For the parameters given above , the average expression at steady state is given by kr fon/γ = 25 . However , rescaling the transition rates does change the shape of the distribution as shown in Fig 2 ( b ) –2 ( d ) [20] . When switching is slow , the gene stays in the “on” and “off” states long enough to observe individual high and low peaks corresponding to the “on” and “off” states , as in shown in Fig 2 ( b ) . However , for intermediate switching rates , the gene does not spend enough time in the “off” state for bursts to decay or enough time in the “on” state for large populations to accumulate ( see Fig 2 ( c ) ) . At fast switching rates the “on” and “off” states come to a fast quasi-equilibrium , and the time-averaged system approaches a Poisson process , where the effective production rate is kr fon . For the bursting gene expression model , all moments of the distributions can be computed exactly from Eq 35 in the Materials and Methods section , even when the RNA distributions are highly non-Gaussian [42] . Since the previous example has already verified the accuracy of the FSP-FIM when the expression has a Poisson distribution , we now verify the FSP-FIM for the slow switching case in which the distribution is bimodal ( α = 0 . 1 ) . To our knowledge an exact FIM solution is not known for the bursting gene expression model , so we evaluate the different FIM approximations by finding the sampling distribution of the MLE , and we compare the covariance of this distribution to the inverse of the FIM [11] . To do this , we sample from p ( X; t , θ* ) under reference parameter set θ* to generate 200 simulated data sets , each with independent RNA measurements for 1 , 000 cells . We then allow koff and kr to be free parameters , and we find θ ^ for each of the 200 data sets . Fig 3 compares the 95% confidence intervals found by taking the inverse of the FIM and through MLE estimation using simulated data for the FSP likelihood ( Eq 6 ) shown in Fig 3 ( a ) , the LNA-based likelihood ( Eq 36 in the Methods section ) shown in Fig 3 ( b ) , and the SM-based likelihood ( Eq 36 in the Methods section , Supplementary Eq . 10 ) shown in Fig 3 ( c ) . Fig 3 ( a ) shows that the CRB predicted by the FSP-FIM matches almost perfectly to the confidence intervals determined by MLE analysis of independent data sets . S3 Fig ( left column ) shows that this estimate is consistently accurate over multiple different experiment designs . In contrast , the LNA-FIM dramatically overestimates the information and predicts confidence intervals that are much smaller than are actually possible ( Fig 3 ( b ) and S3 Fig , center column ) . The SM-FIM does a better job than the LNA in that it matches the MLE analysis for some experimental conditions ( Fig 3 ( c ) ) but much less well for other conditions ( S3 Fig , right column ) . We note that the three different FIM estimates yield different principle directions and different magnitudes for parameter uncertainty ( Fig 3 ( d ) ) , but in all cases the FSP-MLE matches to the FSP-FIM and results in the tightest MLE estimation . Having verified the FSP-FIM for the bursting gene expression model with multiple parameter sets , we next explore how the information changes as a function of the system parameters . Fig 4 shows the determinant of the FIM ( also known as the D-optimality or information density ) for the bursting gene expression model as a function of the switch rate scaling factor , α , using the LNA-FIM ( purple ) , SM-FIM ( green ) and FSP-FIM ( blue ) approximations . In the limit of fast switching ( i . e . α → ∞ ) , the expected information converges to nearly the same value for all approaches , as expected for a Poisson distribution with high effective population size of λ = 25 RNA . However , in the non-Gaussian regimes with slow switch rates , the LNA-FIM over-estimates and SM-FIM under-estimates the information compared to the verified FSP-FIM approach . We note that these differences arise despite the fact that the bursting gene expression model has linear propensity functions , which allows for closed and exact computation of the statistical moments . Next , having verified the FSP-FIM for its ability to accurately estimate the FIM for different parameter sets , we explore the use of the FSP-FIM to design experiments that maximize information . Specifically , we will use classical FIM-based experiment design approaches to choose single-cell experiments first for the bursting gene expression model above , and then for a nonlinear toggle model for which moments can no longer be computed exactly . We consider two different metrics of the FIM , which are frequently used in model-driven experiment design [9 , 12] . The first of these is E-optimality ( presented in the main figures ) , which corresponds to the smallest eigenvalue of the FIM . By finding the experiment which maximizes this eigenvalue , the information is increased in the principle direction of parameter space in which the least information is known ( i . e . the parameter uncertainty is highest ) . The second FIM criteria is D-optimality ( presented in supplemental figures ) , which corresponds to the determinant of the FIM . By maximizing the determinant of the FIM over the experiment design space , one finds an experiment which minimizes the volume of the uncertainty in parameter space . We note that many other experimental design criteria are possible , and the choice of criteria depends on what one desires to learn about the system .
Fluctuations in biological systems complicate modeling by introducing substantial variability in gene expression among individual cells within a homogeneous population . This variability contains valuable and quantifiable insights [20] , but data with multiple peaks and long tails , such as those collected using smFISH , are often poorly described by modeling approaches that only make use of low-order moments of such distributions [26] . The FSP approach [27] has previously been used to identify and predict gene expression dynamics for complex and rich single-molecule , single-cell data [2 , 29 , 30] . In this work , we have developed the FSP-based Fisher information matrix , which extends the FSP analysis to allow rigorous design of experiments that are optimally informative about the model’s parameters . The FSP-FIM uses a novel sensitivity analysis , which requires solving a system of ODEs that is twice the size of the FSP dimension for each parameter , and therefore should be computationally tractable for any problem to which the FSP can be applied . The local sensitivity of each parameter is independent of the other parameters , so the computation is easily parallelized among multiple processors . We verified that the FSP-FIM approach matches the information for the constitutive gene expression model , whose response follows a Poisson distribution ( Fig 1 ) , and for which the FIM can be computed exactly . The FSP-FIM also matches to classical FIM approaches that assume normally distributed data ( LNA-FIM ) or very large data sets ( SM-FIM ) in the limiting case when the data distributions are close to being Gaussian ( Figs 1–4 ) . For systems where data is not Gaussian and for which there is no exact FIM formula , we showed that the FSP-FIM is more accurate than traditional approaches ( Figs 4 and 5 ) , which we validated by generating many independent data sets and comparing the inverse of the FSP-FIM to the variance in the MLE estimates ( Figs 3 and 6 ) . We showed that the choice of FIM analysis can lead to different optimal experiment designs ( Fig 5 ) . For example , Fig 5 and S3 Fig show that the LNA-FIM can substantially overestimate the information of certain experiments compared to the full , correct information obtain using the FSP-FIM , which could mislead researchers to choose experiment designs that are much worse than they expect . In practice , overestimation of the Fisher information can have the further deleterious effect of overconfidence in poor parameter estimates , which can result in model bias and poor predictions as we observed recently in [26] . Furthermore , the LNA-FIM is not self-consistent , and often overestimates the information even compared to the ellipse found from sampling the MLE with the Gaussian likelihood function . On the other hand , we found that the SM-FIM under-estimated the information for the bursting gene model , but the amount of underestimation varied substantially with experimental conditions , which could cause researchers to reject otherwise informative experiments . In contrast to these moment-based approaches , the MLE sampling using the FSP approach always provided the best parameter estimates ( Fig 3 and S3 Fig ) , and the FSP-FIM was always consistent with the confidence intervals verified by sampling ( Figs 1 , 3 and 5 , S1–S3 Figs ) , even for the case of nonlinear reaction propensities for which exact moments cannot be found ( Fig 6 ( a ) , and S4 Fig ) . In our analysis of the non-linear toggle model , we allowed for the independent control of three experimental variables ( Fig 7a ) , and found experiments that optimize particular criteria of the FIM . Furthermore , we showed that other experiments very near to the optimal experiment in the design space can be significantly less informative than the optimal experiment ( Fig 7 ( b ) –7 ( e ) and S6 Fig ) . Choosing between such similar experiment designs is non-trivial and would be difficult or impossible using intuition alone . On the other hand , we explored the effects of parameter uncertainty on FSP-FIM-based experiment design , and we found that parameter rankings are relatively robust to parameter uncertainty , even when the absolute value of the FSP-FIM is sensitive ( Fig 7 ) . We found that that the choice of optimal experiments depends on the number of experiments to be completed ( Table 2 ) . For example , the optimal set of two experiments may not contain the optimal single experiment . Sometimes , the FIM for a given experiment may be singular or nearly singular , indicating that the model under investigation is unidentifiable for the current parameterization and experiment design . In such a case , the FIM-eigenvectors corresponding to the near-zero eigenvalues indicate specific linear combinations of parameters that are poorly constrained ( e . g . , ‘sloppy’ directions [47] ) . If a second complementary experiment can shift the orientation of these sloppy vectors , then those parameters may yet be uncovered through combinations of multiple experiments . Alternatively , if a given combination of parameters remains unidentifiable for all admissible experiments , then these irrevocably sloppy directions may be used to reformulate the model into one that has a reduced set of fully identifiable parameters . We note that as one conducts new experiments and collects new data , parameter posteriors will need to be updated . As this occurs , optimal experiments may also need to be adjusted ( e . g . , through application of a Bayesian experiment design framework [48] ) , and future developments are needed to incorporate FSP-FIM computations within such iterative frameworks . Our results show that the FSP-FIM performs better than previous approaches for gene regulation models with low molecule counts or nonlinear reaction rates . Previous studies have demonstrated many realistic systems for which such FSP can be used to identify and predict stochastic dynamics in numerous biological systems [2 , 6 , 19 , 26 , 29–32 , 49] . Each of these studies has used different experimental input signals , such as temporal salinity profiles [2 , 26] , temperature [29] , or chemical induction [19 , 30] . Modern optogenetic experiments promise to allow for even more robust and flexible control of input signals to control cellular behavior [7 , 50 , 51] . For such studies , the FSP-FIM could now be used to optimize these signals to achieve maximally informative experiments . Like any other tool , the FSP-FIM also has its associated challenges . Our initial investigations focused on intrinsic stochastic fluctuations of small biochemical processes , and we used simulated data to verify our new computational tools . For models with large molecular counts of four or more species or with the addition of mechanisms to account for extrinsic variability , existing methods to solve the FSP-FIM will remain intractable until more efficient probability density based CME analyses can be developed to address such problems [52–56] . Until higher dimension CME approaches are developed , approximate moment-based experiment design methods , such as the SM-FIM and LNA-FIM , may remain the only available options to design experiments for large biochemical pathways . We also note that real experiments come with additional sources of noise , such as the errors or uncertainties associated with experimental measurements . For example , in smFISH data analysis , image processing settings give rise to variability in final RNA counts due to density dependent co-localization of RNA molecules . This measurement uncertainty may have a non-negligible effect on parameter inference , and future controlled experiments are needed to elucidate the degree to which such effects depend on optical imaging settings . Fortunately , such variabilities are easily incorporated within the framework of the FSP analysis . For example , previous work has used a simple linear transformation to adapt FSP analyses to include the effects of noisy GFP fluorescence emission and background autofluorescence when comparing integer-valued biochemical models to flow cytometry data in arbitrary continuous units of fluorescence [19] . Once adapted to take these transformations into account , the FSP-FIM could be used to design experiments to minimize the effects of measurement noise . New experimental capabilities are creating an enormous potential to probe single-cell biological responses . These capabilities are making it ever more difficult to choose what species in the system to measure , whether to measure joint distributions ( i . e . measure the RNA counts from multiple genes in the same cells ) or marginal distributions ( only measure RNA counts from a single gene at a time ) , or in what condition . Furthermore , different experiments have different costs , and the experimentalists must not only optimize their information about model parameters , but also consider the trade-off between collecting more data and the cost of a given experiment . By providing a new computational tool to iteratively improve models and design experiments for an important class of biological problems , the FSP-FIM will help to improve quantitative predictive modeling of gene expression .
The change of probability p ( xl ) with respect to small changes in parameter θj describes the sensitivity of the lth state in the Markov process to the jth parameter [33 , 57] . These local sensitivities can be calculated by transforming the linear ODEs describing the time evolution of the probabilities of the state space d d t p ( t ) = f ( p ( t ) , θ , t ) into a set of ODEs describing the time evolution of the sensitivities . Given an initial condition , the solution to the CME is: p ( t ; θ ) = p ( t 0 ) + ∫ t 0 t f ( p ( s ; θ ) , θ , s ) d s ( 29 ) Taking partial derivatives with respect to θ , ∇ θ p ( t ; θ ) = ∫ t 0 t [ ∇ θ f ( p ( s ; θ ) , θ , s ) + ∇ p f ( p ( s ; θ ) , θ , s ) ∇ θ p ( s ; θ ) ] d s . ( 30 ) We can now describe the sensitivities S ≡ ∇θ p as they evolve with time , by taking the time derivative of the equation above . For the FSP , the right-hand side f ( p ( t;θ ) , θ , t ) = A ( θ , t ) p ( t ) , and ∇ θ f ( t , p ( t ; θ ) , θ ) = ( ∇ θ A ( θ ) ) p ( t ) ( 31 ) ∇ p f ( t , p ( t ; θ ) , θ ) = A ( θ ) ( 32 ) In many cases , including all models formulated using mass-action kinetics , the generator A can be written as a linear combination of the model parameters , i . e . A = ∑θi Bi , and the derivative with respect to the ith parameter can be found , ∂ ∂ θ i A = ∂ ∂ θ i ( θ i B i ) = B i . ( 33 ) Using this notation , Eq 30 is reduced to the set of linear ODEs for each parameter θi , d d t ( p ( t ) S i ( t ) ) = ( A 0 B i A ) ( p ( t ) S i ( t ) ) . ( 34 ) In practice , Eq 34 can be computed in parallel for each parameter , and the computation of sensitivities for K parameters is equivalent to solving K sparse systems of ODEs , each twice the size of the FSP computation . Current state-of-the-art approaches for single-cell , single-molecule experiment design rely on computing moments of the CME . Such statistical moments may be computed exactly for systems with affine-linear propensities [42] . The uncentered moments of the CME , E { x m } , where m = [ m 1 , m 2 , … , m N s ] is a vector of integers corresponding to the m i th power of the i th species in x , and the entire moment xm is found according to the following formula: E { x m } d t = E { ∑ j = 1 M w j ( x ) [ ∏ i = 1 N ( η i + Ψ i j ) - ∏ i = 1 N η i m i ] } . ( 35 ) In the limit of large numbers of molecules reacting in a well-mixed solution , the linear noise approximation ( LNA ) may be applied to CME [25] . In such cases , molecule numbers are considered to be Gaussian , and the well-known Gaussian form of the FIM may be applied [8] . If the observed data is assumed to come from a multivariate Gaussian distribution with means μ ( t ; θ ) = [ μ 1 ( t ; θ ) , μ 2 ( t ; θ ) , … μ N s ( t ; θ ) ] T and covariance matrix Σ ( t;θ ) , such as those in Eq 35 , the likelihood is given by: L ( D ; μ , Σ ) = ∏ t = t 1 t N t ∏ i = 1 N c ( 2 π N o | Σ ( t ) | ) - 1 2 × exp ( - 1 2 ( d i ( t ) - μ ( t ) ) T Σ - 1 ( t ) ( d i ( t ) - μ ( t ) ) ) ( 36 ) and the FIM is well-known [10 , 11] F I M i , j = ∂ μ ∂ θ i T Σ - 1 ∂ μ ∂ θ j + 1 2 t r a c e ( Σ - 1 ∂ Σ ∂ θ i Σ - 1 ∂ Σ ∂ θ j ) . ( 37 ) Another approach , developed in [9] is to use a likelihood function that takes the sample mean and sample variance to be jointly Gaussian , and thus requires the computation of up to the 4th moments in Eq 35 . In the supplement , we reproduce the formulae from [9] relevant to this study .
|
A main objective of quantitative modeling is to predict the behaviors of complex systems under varying conditions . In a biological context , stochastic fluctuations in expression levels among isogenic cell populations have required modeling efforts to incorporate and even rely upon stochasticity . At the same time , new experimental variables such as chemical induction and optogenetic control have created vast opportunities to probe and understand gene expression , even at single-molecule and single-cell precision . With many possible measurements or perturbations to choose from , researchers require sophisticated approaches to choose which experiment to perform next . In this work , we provide a new tool , the finite state projection based Fisher information matrix ( FSP-FIM ) , which considers all cell-to-cell fluctuations measured in modern data sets , and can design optimal experiments under these conditions . Unlike previous approaches , the FSP-FIM does not make any assumptions about the shape of the distribution being measured . This new tool will allow experimentalists to optimally perturb systems to learn as much as possible about single-cell processes with a minimum of experimental cost or effort .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"ultraviolet",
"radiation",
"experimental",
"design",
"random",
"variables",
"light",
"covariance",
"electromagnetic",
"radiation",
"simulation",
"and",
"modeling",
"research",
"design",
"probability",
"distribution",
"mathematics",
"algebra",
"research",
"and",
"analysis",
"methods",
"gene",
"expression",
"probability",
"theory",
"approximation",
"methods",
"physics",
"linear",
"algebra",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"eigenvalues"
] |
2019
|
The finite state projection based Fisher information matrix approach to estimate information and optimize single-cell experiments
|
Intrinsic disorder is more abundant in eukaryotic than prokaryotic proteins . Methods predicting intrinsic disorder are based on the amino acid sequence of a protein . Therefore , there must exist an underlying difference in the sequences between eukaryotic and prokaryotic proteins causing the ( predicted ) difference in intrinsic disorder . By comparing proteins , from complete eukaryotic and prokaryotic proteomes , we show that the difference in intrinsic disorder emerges from the linker regions connecting Pfam domains . Eukaryotic proteins have more extended linker regions , and in addition , the eukaryotic linkers are significantly more disordered , 38% vs . 12-16% disordered residues . Next , we examined the underlying reason for the increase in disorder in eukaryotic linkers , and we found that the changes in abundance of only three amino acids cause the increase . Eukaryotic proteins contain 8 . 6% serine; while prokaryotic proteins have 6 . 5% , eukaryotic proteins also contain 5 . 4% proline and 5 . 3% isoleucine compared with 4 . 0% proline and ≈ 7 . 5% isoleucine in the prokaryotes . All these three differences contribute to the increased disorder in eukaryotic proteins . It is tempting to speculate that the increase in serine frequencies in eukaryotes is related to regulation by kinases , but direct evidence for this is lacking . The differences are observed in all phyla , protein families , structural regions and type of protein but are most pronounced in disordered and linker regions . The observation that differences in the abundance of three amino acids cause the difference in disorder between eukaryotic and prokaryotic proteins raises the question: Are amino acid frequencies different in eukaryotic linkers because the linkers are more disordered or do the differences cause the increased disorder ?
Eukaryotic cells are more complex than prokaryotic cells , and therefore , have an increased need for regulation . They also contain organelles , have more complex genes and a more advanced chaperonin system enabling the folding of longer proteins [1] . In response to the increased complexity , eukaryotic proteomes have evolved to differ significantly from prokaryotic proteomes . The most notable differences are that; ( i ) eukaryotic proteins are longer [2–5] , ( ii ) multi-domain proteins are more abundant in eukaryotes [6–8] , ( iii ) domain repeats are frequent in multicellular organisms [9] , and ( iv ) eukaryotic proteins have a higher fraction of disordered residues [10] . The increased length of eukaryotic proteins is , at least partly , a consequence of them containing more domains [11] . With more multi-domain proteins , it follows that eukaryotic proteins have more linker regions—connecting the domains [12] . Further , the increased number of domain repeats appears to be a unique feature of multicellular organisms [9] . These repeats have been proposed to provide eukaryotes with an additional source of variability to compensate for low generation rates [13] and are important for signalling . The origin of the increase in intrinsic disorder in eukaryotic proteins is less well understood . Intrinsic disorder is frequent in all eukaryotic phyla , and even among viral proteins [14] . In earlier studies , about 10% of the residues in prokaryotes are predicted to be disordered compared with 30% in eukaryotes [15–18] . Disordered regions are over-represented in regulatory proteins [19] , providing a possible explanation for the increase of intrinsic disorder in eukaryotes . Ahrens et al . proposed that the increased intrinsic disorder in eukaryotic is a result of lower selective pressure due to the smaller effective population size in eukaryotes [15] . The observation that ancient eukaryotic genes are less disordered than young or random genes [20] supports this . However , a large number of functionally important intrinsically disordered regions have been described [21 , 22] . Functions associated with disordered regions include; to present short linear motifs that are important for binding [19] and to enable post-translational modification that preferentially occurs in intrinsically disordered regions [23 , 24] . Likely , at least some of the intrinsically disordered regions in eukaryotic proteins are functionally important . The vast majority of studies of intrinsic disorder are based on predictions [25] and although the best predictors use multiple sequence alignments [26] , even simple predictors that only use the amino acid sequence identify the difference between eukaryotes and prokaryotes [27] . The average “disorder propensity” , as measured by the TOP-IDP scale [28] , is also significantly higher for eukaryotic proteins than for prokaryotic proteins . Polar and charged amino acids , together with proline , are the most disorder-promoting residues . Thus , proteins with a higher fraction of these residues are ( predicted to be ) more disordered . Therefore , there should be an increase in the abundance of these amino acids in eukaryotic proteins or a decrease of the order promoting residues . However , to the best of our knowledge , shifts of amino acid frequencies between eukaryotic and prokaryotic proteins have not earlier been used to analyse the difference in intrinsic disorder . Over evolutionary times there exist many possibilities for amino acids to change in a protein family without the loss of function [29] . Most protein families contain members that have less than 20% sequence identities [30]; i . e . for most proteins , it is possible to replace more than 80% of the residues and still maintain its function . Further , protein design experiments have shown that it is possible to design functional proteins with a limited [31] , or biased [32] , set of amino acids . Therefore , an organism should be able to adapt its amino acid frequencies if an advantage to do so exists . Multiple factors can affect systematic shifts of amino acids frequencies , and one of the most notable is the GC content of the genome . Amino acids encoded by high GC codons are enriched in high GC genomes and vice versa . This trend is particularly strong among recently created genes but also exists for ancient genes [20] . It has been shown that amino acids with codons enriched in GC are disorder-promoting [33] , explaining why de novo created proteins in yeast ( low GC ) appear to be ordered while in Drosophila ( high GC ) such proteins are predicted to be disordered [20] . The general trend of amino acid gains and losses has also been studied before , and it has been proposed that the amino acids ( except serine ) that appeared to increase in frequency recently were not incorporated in the first genetic code [34] . However , the statistical methodology used in that study has been questioned [35] . Further , it has been observed that the frequency of tyrosine has decreased in Metazoans compared to yeast [36] , and histidine and serine frequencies increase from high-temperature thermophiles to prokaryotic mesophiles and further to eukaryotes while valine shows the opposite trend [37] . Finally , a trend of increasing polar amino acids in eukaryotes has been reported [38] . Some of these changes can contribute to the increased disorder in eukaryotes , but until now , studies of intrinsic disorder have not taken shifts of amino acid frequencies into account . In this study , we try to identify the molecular properties that underlie the difference in intrinsic disorder between eukaryotes and prokaryotes . First , we show that the difference in disorder can primarily be attributed to that linker regions in eukaryotes are , not only more abundant but also more disordered . Next , we show that differences in serine , proline , and isoleucine frequencies can explain the difference in intrinsic disorder between eukaryotic and prokaryotic linkers .
The dataset used in this study originates from the complete bacterial , archaeal and eukaryotic proteomes in UniProt [39] as of December 2017 . However , differences in GC composition complicate the comparison of amino acid distributions as the frequency of some amino acids is strongly dependent on the GC content of the genome , S1 Fig . In the prokaryotic kingdoms , there exist a significant fraction of genomes with high GC content , S2 Fig . We tried several methods to compensate for differences in amino acid frequencies caused by the differences in GC . One possibility is to use an ANOVA test , S1 Table . The general conclusions are similar using any of these methods , but if GC is completely ignored significant differences can be missed . After several tries , we do believe that the easiest way to compensate for GC is to ignore all genomes with extreme GC content . In addition to the simplicity , this removal also makes it possible to compare trends within protein families without compensating for the GC content . Therefore , we excluded all genomes with a GC content of more than 60% or less than 20% . The resulting set of genomes have a similar GC content in all three kingdoms , and the average GC is 43-44% with a standard deviation of 8% , S2b Fig . All genomes from Mycoplasma , Spiroplasma , Ureaplasma , and Mesoplasma were also ignored as they have another codon usage—which influences the expected amino acid frequencies . The final dataset contains 26 , 274 , 724 protein sequences from 6 , 373 genomes , divided into 4 , 905 bacterial , 308 archaeal , and 975 eukaryotic . Different numbers of proteins of a particular type or differences within proteins of the same type can cause differences at the proteome level . To distinguish these scenarios , we divide the complete proteomes into subsets using Pfam [40 , 41] . First , we identified 4 , 165 shared Pfam domains that are present in at least ten eukaryotes and ten prokaryotes , and where none of the kingdoms makes up of more than 99 . 9% of all the members . 1 , 764 of these domains are present in all three kingdoms . We define a set of “shared proteins” as all proteins that contain at least one of these “shared domains” . Proteins that only contain Pfam domains that are unique to one of the kingdoms are referred to as ( kingdom ) “specific proteins” , and proteins without any Pfam domains are called “no domain” proteins , see Fig 1 . Also , within one group of proteins , proteome-wide differences might be caused by the abundance of different regions or differences within similar regions . Therefore , we divided the “shared proteins” further into regions , see Fig 1 . Regions corresponding to any of the 4 , 165 Pfam domains , that exist both in prokaryotes and eukaryotes , are called “shared domains” , while regions assigned to any other Pfam domain are “specific domains” , and all regions that are not assigned to a Pfam domain are classified as “linker regions” . The linker regions plus the no-domain proteins should be similar but not identical to the dark proteome [42] . For each of these six groups , we analysed length , disorder , amino acid frequencies , and other properties independently . As shown in Fig 1 , a protein can contain zero , one or multiple regions of a particular type . Therefore , if a protein contains two “shared domains” the length of shared domains in that protein is the sum of the length of both the domains . The processed datasets , as well as all scripts , are available from https://figshare . com/articles/Dataset_for_paper/7478381 . For each protein , we estimated the intrinsic disorder using two tools: IUPred [27] and TOP-IDP [28] . IUPred exploits the idea that in disordered regions , amino acid residues form less energetically favourable contacts than residues in ordered regions . IUPred does not rely on any external information besides the amino acid sequence and is therefore extremely fast and suitable to predict disorder for large data sets . We used the recommended cut-off and assigned a residue to be disordered if its IUPred value is higher than 0 . 4 [43 , 44] . We report the results using IUPred long disordered predictions . Using the short version of IUPred or a different cut-off gives almost identical results , S2–S4 Tables . We also calculate the average disorder propensity using the TOP-IDP scale [28] for each region . Properties , including length , amino acid type and disorder were analysed independently for each protein region , as described in Fig 1 . Comparisons were performed between regions of different types and between kingdoms . Statistical significances were calculated using Students T-tests , but the numerous data points make even small differences statistically significant . For instance , the predicted number of disordered residues among the shared domains is small ( 21 . 3 in bacteria vs . 27 . 1 in eukaryotes ) , S2 and S3 Tables , but significant ( P < 1 . 3 * 10−8 ) . For many other comparisons , the P-values are smaller than 10−200 . Therefore , we do not believe it is of relevance to report each P-value for all comparisons . Instead , we have just included the standard errors in relevant figures and S2–S4 Tables .
As shown before [5 , 8 , 45] , eukaryotic proteins are on average longer than prokaryotic proteins see Fig 2 and S2–S4 Tables . The group of proteins with “shared domains” is longer than proteins with only “specific domains” , and the proteins without domains are even shorter . However , in all three groups , eukaryotic proteins are significantly longer than the prokaryotic proteins . We have , in an earlier study , contributed the difference in length to that eukaryotic proteomes contain more multi-domain proteins [5] . In that study , we assumed that long linker regions contained missed domains , and this contributed to the assumption that the increase in multi-domain proteins was a driver for the difference in length between eukaryotic and prokaryotic proteins . However , given the insights from studies of disordered regions [12] and the dark proteome [42] , it is now clear that long linker regions do not necessarily contain unassigned domains . Therefore , we do not assign domains to long unassigned regions . To understand the origin of the difference in length between eukaryotic and prokaryotic proteins , we choose to study the shared proteins in more details . Among the 14 million proteins with “shared domains” the average length of the eukaryotic proteins is 532 vs . 345 for bacterial protein and 309 for proteins in Archaea . The number of residues in “shared domains” is roughly equal in the three kingdoms , 218 to 233 , and the average number of residues assigned to “kingdom specific domains” is , although higher in eukaryotes , quite low ( 27 in bacteria , 19 in Archaea , and 49 in eukaryotes ) , see Fig 3 . In contrast , the number of residues in “linker regions” differs significantly between the kingdoms , in eukaryotes , 48% of all residues are assigned to “linker regions” , compared to only 31% in prokaryotes , S2–S4 Tables . Thus , the length of “linker regions” comprises > 80% of the length difference between eukaryotic and prokaryotic proteins . Eukaryotic proteins have more residues assigned to linker regions . Linkers can be located at one of the termini or between two domains . In all three kingdoms , each of the termini contains roughly 40% of the linker residues , and linkers between domains ( central ) the remaining 20% , S2–S4 Tables . Independent on location , linkers are more than twice as long in eukaryotes than in prokaryotes . To understand how the linkers differ between eukaryotes and prokaryotes , it is necessary also to consider differences between eukaryotic and prokaryotic domains . Many Pfam domains only cover the central most conserved core of a domain and not variable regions at the termini [46] . Eukaryotic domains are known to show increased variability , possibly contributing to the extended linker regions [47] . Therefore , it is not impossible that extensions of existing domains cause some of the increased linker lengths in eukaryotes . However , we do believe that these additional residues should not be significantly more ordered than other residues within the domains . Therefore , variations within domains should not be the principal cause for the increased disorder in eukaryotic proteins . Next , we studied the disorder in the different groups of proteins . All three groups of eukaryotic proteins are more disordered than prokaryotic ones , see Fig 2 . In agreement with earlier studies [15 , 18 , 48 , 49] , 12% of the residues in prokaryotes are predicted to be disordered compared with 32% in eukaryotes , S2–S4 Tables . Proteins that are unique to eukaryotes are more disordered than those that contain “shared domains” , and eukaryotic proteins without any Pfam domains are the most disordered with 42% disordered residues . The observation that proteins unique to eukaryotes have increased disorder supports the earlier observations that young eukaryotic proteins are more disordered than older proteins [20] . For prokaryotic proteins , the disorder content in all three groups of proteins is lower ( 11-15% ) . To understand the origin of the difference in disorder better , we studied disorder in the different regions of the proteins that contain a “shared domain” . First , it can be seen that eukaryotic “specific domains” are more disordered than all other types of prokaryotic or eukaryotic domains , 17% vs . 8-12% , S2–S4 Tables . However , the most significant difference is that eukaryotic “linker regions” are much more disordered ( 38% ) than prokaryotic “linker regions” ( 12-16% ) , see Fig 3 . The difference in disorder can therefore not only be contributed to that “linker regions” are more abundant in eukaryotic proteins , but also to that eukaryotic linkers contain a higher fraction of disordered residues . Eukaryotic and prokaryotic proteins differ both in lengths of different regions and in disorder content . Therefore , it might be of interest to describe an average eukaryotic and prokaryotic protein . The average eukaryotic protein is 450 residues long and contains 32% disordered residues , while an average prokaryotic protein is 300 residues long and contains 12% disordered residues , which infers that the average eukaryotic protein contains 145 disordered residues compared with 32-37 for the prokaryotic proteins , see Fig 4 . Next , eukaryotic proteins have much longer linker regions with 258 vs ≈ 110 residues in prokaryotes , and the eukaryotic linker regions are more disordered , see Fig 3 . 100 of the disordered residues in eukaryotic proteins are located in the linker regions , while prokaryotic linker regions only contain 12-18 disordered residues , S2–S4 Tables . The number of disordered residues within the domains is higher in eukaryotic proteins , 36 vs 17-24 . Anyhow , this demonstrates that the increase in disorder is dominated by the increase in disorder within the linkers . Above , we show that eukaryotic proteins are more disordered than prokaryotic proteins because their “linker regions” are both longer and more disordered . However , ( predicted ) intrinsic disorder is primarily caused by differences in amino acid frequencies . Therefore , we studied the difference in amino acid frequencies between eukaryotic and prokaryotic proteins . One way to compare properties of different regions is to compare the amino acid distributions in the entire regions and then cluster the regions , see Fig 5 . In the heat map , the most substantial difference between regions is that the amino acid frequencies of eukaryotic linkers are distinct from all other regions . It can also be observed that all regions in Archaea cluster together , while the eukaryotic domains and all bacterial regions form the third cluster . However , this difference is much smaller . To understand what causes the eukaryotic linkers to have unique amino acid distributions , we compared the amino acid frequencies between eukaryotic and prokaryotic regions , see Fig 6 and Table 1 . Here , it can be seen that there exist three amino acid , isoleucine , serine and proline , whose frequencies differ by more than 1 . 5% between eukaryotic “linker regions” and linkers in either of the prokaryotes . These differences are also notable in the heat map , see Fig 5 . Further , the frequencies of these amino acids also differ within the shared domains , but to a smaller degree , see Fig 6b . Finally , a two-way ANOVA test shows that isoleucine , proline and serine are the amino acids with the most significant differences between the eukaryotic and bacterial proteins when including the GC content , S1 Table . It should be noted that the shifts of isoleucine and proline are small if the GC content of the genomes is ignored . However , the increase in serine frequency among eukaryotes is easy to detect , and it is a surprise to us that this has not been highlighted before . The amino acid frequency in different regions shows that not all disorder-promoting amino acids increase in frequency in eukaryotic linkers . The difference in disorder is instead caused by the shift in frequencies of only three amino acids , isoleucine , serine , and proline . All three amino acids contribute to the increased disorder in eukaryotic linkers , and if these three amino acids are ignored , there is no significant difference in disorder propensity between eukaryotes and prokaryotes , S3 Fig . However , it is not clear if the increased disorder in eukaryotic linkers is primarily a consequence of changes in amino acids frequencies , or if the need for increased disorder drives the changes in amino acid frequencies—a chicken and egg problem . Eukaryotic proteomes are in general larger than prokaryotic proteomes; this is partly due to an expansion of protein families by gene duplications . For functional reasons , different protein families have different amino acid distributions . Therefore , it is possible that the differences in the amino acid frequency that we observe when studying an entire proteome are due to the different frequencies of protein families . However , to better under the origin of the amino acid frequency differences , we examined the amino acid frequency of all shared Pfam domains independently . The reason to study domains and not the linkers is that the linkers are challenging to align and differ significantly in length , while the domains are of similar length and already aligned in Pfam . Further , the serine and isoleucine differences are also present among the shared domains , see Fig 6b . In Fig 7 , the differences between the amino acid frequencies in the prokaryotic domains are compared with the amino acid frequencies in the corresponding eukaryotic domains . Only Pfam families with at least 100 members among both bacteria and eukaryotes are included to avoid statistical outliers ( Archaea was ignored in this filtering ) . In 84% of the families , the eukaryotic members have more serine , in 80% fewer isoleucine and 70% more proline , i . e . the shifts in frequencies are observed in a majority of the families . We also tried to identify any trends among the families with extreme amino acid frequencies , both by examining individual families and by mapping to GO-slim terms , using pfam2go [50–52] . The GO terms with the most substantial differences in amino acid frequencies are listed in S5 Table . However , to the best of our ability , we cannot identify any particular functional subset of proteins where the difference in frequency significantly differs from the general picture . Therefore , the differences in frequencies do not appear to be caused by shifts in the frequency of some particular group of proteins . Instead , there seems to exist a systematic shift in the frequencies between eukaryotes and prokaryotes present in most protein families . A difference between eukaryotes and prokaryotes is that eukaryotic cells have organelles . The amino acid content of proteins in different organelles differs; therefore , it would not be implausible that the different amino acid frequencies could be affected by the compartmentalization of the eukaryotic cell . However , in all membrane and non-membrane parts of all organelles , the frequencies of serine and proline are higher in eukaryotes than in prokaryotes , see S6 Table . Further , in all but three organelles , the isoleucine frequency is lower in eukaryotes . Some bacteria within Planctomycetes , Verrucomicrobiae , and Chlamydiae have quite complex membranes , possibly indicating primitive organelles [53] . However , all these phyla have bacterial levels of serine , proline and isoleucine , see Fig 8 . Therefore , the compartmentalization of the eukaryotic cell does not appear to explain the differences in amino acid frequencies between eukaryotes and prokaryotes . Within protein regions , there exist different structural elements , such as helices , sheets loops and disordered regions . Amino acids have different preferences for different structural elements . Therefore , to investigate the preferences of amino acids in different structural elements , we compared the amino acid frequencies in different structural regions within the “shared domains” . Here , we only use the Pfam families where there was at least one structure available in PDB , and we assume that the secondary structure is conserved within the entire Pfam family . The reason to use only the “shared domains” is that the structural information of the linkers is limited . Using the secondary structure annotation , available from Pfam , we then assign each residue into one out of three categories , Helix , Sheet or Coil , using the most frequent annotation in Pfam . Unassigned positions , i . e . , residues corresponding to the parts of the Pfam domains that are not present in any PDB structure , we do refer to as disordered , as often done when training disorder predictors [54] . The amino acid frequencies in each structural region are shown in Fig 9 and S6 Fig . As expected , the serine and proline frequencies are highest in loops and disordered regions . However , when comparing amino acid frequencies between the kingdoms , it can be seen that the serine frequency is increased in all secondary structures in eukaryotes compared with prokaryotes . The most substantial difference is observed in the disordered regions ( 2% ) . For proline and isoleucine , a smaller , but still statistically significant ( P < 10−4 ) can be observed in all secondary structure classes , i . e . he frequency differences of serine , proline and isoleucine are widespread and not unique to a particular protein element . What is the underlying reason for the shifts in amino acid frequencies ? One possible reason for the higher fraction of serine in eukaryotic organisms is that serine , together with threonine , are targets for Ser/Thr kinases [55] . Phosphorylation of serine and threonine is one of the critical regulatory pathways in eukaryotes , but also present in Archaea [56] . Further , phosphorylation frequently occurs in intrinsically disordered sites [57] . Together this makes it intriguing to speculate that serine frequency is higher in eukaryotic linkers because of the increased need for regulation by kinases . Ser/Thr kinases are prevalent in eukaryotes , but also exist in bacteria such as Planctomycetes [58] . The only fully sequenced genome of this phylum ( Planctomycetes bacterium GWA2_40_7 ) has 6 . 1% serine , typical for bacteria . Further , the largest family of Ser/Thr kinases , Pfam family Stk19 ( PF10494 ) , only exists in eukaryotes and Halanaerobiales . The 2783 Halanaerobiales sequences in UniProt [39] contain 5 . 8% serine , also typical for a prokaryote . The bacterial levels of serine in these organisms show that the presence of Ser/Thr kinases is not necessarily causing an increase in serine frequencies . Phosphorylation can occur at three amino acids , serine , threonine , and tyrosine . Threonine and tyrosine frequencies show no increase in eukaryotic “linker regions” , S4 Fig , even when GC is taken into account , S5 Fig . If phosphorylation by kinases is the primary reason for the serine frequency difference between eukaryotes and prokaryotes , why only serine frequency is increased ? It might be due to that about 75% of the known targets for kinases are serine [59] . It might also be related to the fact that serine is a disorder-promoting residue while threonines and tyrosines are not . Although it is tempting to speculate that phosphorylation contributes to the increase of serine in eukaryotes , there exists no direct evidence that regulation by Ser/Thr kinases is the cause of the increased serine frequency . In contrast to serine , we are not aware of any functional roles , of proline and isoleucine , that are kingdom specific , but some proline-rich structural features might be more prevalent in eukaryotes . In addition to being enriched in loops , proline is frequent in repeat proteins [60] , and in particular , PPP and PPG repeats are frequent in multicellular organisms [61] . Proline repeats are also often found in disordered regions that are important for binding in eukaryotic specific proteins such as P53 [62] . Proline is also frequent in “linker regions” connecting domains [63] . As both repeats and multi-domain proteins are more frequent in eukaryotes , these factors might contribute to the increase of proline in eukaryotic proteins . However , as proline is also more frequent within the eukaryotic linker regions , this does not explain the increase in proline . Prokaryotes ( but not eukaryotes ) use a specific purine-rich sequence on the 5’ side to distinguish initiator AUGs from internal ones [64] . The codons for isoleucine contain 44% Adenosine . Therefore , this could potentially contribute to the higher fraction of isoleucine in prokaryotes . However , as the frequency differences between eukaryotes and prokaryotes also exist in C-terminal regions , this cannot be the only explanation for the difference of isoleucine frequency . In addition to functional reasons for the differences in frequencies , the differences could be caused by general trends in the strength of the selective pressure . Such a model would assume that there is a general preference to shift the frequency of an amino acid from what is expected by chance . Functional selection is typically considered to be the dominant force shaping proteome evolution , but secondary effects such as the cost of producing an amino acid or codon usage preferences can also affect the general trend of amino acid frequencies [65] . The population size of eukaryotes is in general smaller than for prokaryotes causing a lower selective pressure . The amount of intrinsic disorder is lower than expected by chance in both eukaryotes and prokaryotes [20] . Therefore , it is possible that the lower selective pressure could explain why eukaryotes contain more disordered residues if these residues are unfavourable [15] . However , this is not always the case as some disorder-promoting residues , such as arginine , are less frequent than expected by chance ( calculated from random nucleotides ) , while others , including lysine , are more frequent , see Fig 10 and S1 Fig . Therefore , it is unlikely that a purifying selection is the only driving force for the observed shifts in amino acid preferences between eukaryotes and prokaryotes . In bacteria , one reason to reduce the frequency of an amino acid is the energetic cost to produce it [66] . Serine is among the least costly amino acids to make both aerobically and anaerobically [66 , 67] , S7 Table . Proline is cheaper than most amino acids to make , while isoleucine is among the most expensive ones . Therefore , the cost of producing amino acids would predict that serine and proline frequencies decreased in the species with higher selective pressure , i . e . the prokaryotes and isoleucine increased , opposite to what is observed . It has also been reported that high serine levels are toxic [68 , 69] , possibly contributing to reduced serine levels are reduced in prokaryotes . Anyhow , none of the explanations discussed above can fully explain the shift in frequency for all three amino acids . Further , if there just was a selective pressure to decrease the amount of disorder , it is not clear why only the frequencies of three amino acids should be affected . Therefore , it is unlikely that the reduced selective pressure in eukaryotes can explain the shifts in amino acid frequencies . Here , we confirm earlier observations that eukaryotic proteins are more disordered than prokaryotic proteins . We show that more extended and more disordered linkers cause a systematic increase in intrinsic disorder in eukaryotic proteins . Further , we show that the increased disorder in the linkers originates from a systematic shift in the frequency of only three amino acids , serine , proline , and isoleucine . Serine and proline are more frequent in eukaryotic proteins than in prokaryotic proteins , while isoleucine is less frequent . For serine , the difference holds for all phyla , protein families , structural regions of proteins and type of protein but is most pronounced in disordered and linker regions . The proline and isoleucine differences are also observed in most classes of proteins and regions but are affected by differences in GC levels of the genomes . Anyhow , it is safe to assume that the differences in amino acid frequencies occurred soon after the three kingdoms split and have been maintained during the last billion years . It is not clear if the increases of serine and proline and decrease in isoleucine cause the increased disorder in eukaryotic proteins , or are a consequence of it . It is tempting to speculate that the increase in serine is related to its importance as a target for regulatory kinases , but direct evidence for this is lacking . Further , the increased need for regulation in eukaryotes does not explain the shift in proline and isoleucine frequencies . Anyhow , the observation that not all disorder-promoting amino acids are increased in eukaryotic linkers makes it clear that earlier explanations of the increased disorder in eukaryotic proteins are too simplified . Further , why just isoleucine , serine and proline frequencies differ between eukaryotes and prokaryotes remains an open question that requires further analysis .
|
Intrinsic disorder is essential for various functions in eukaryotic cells and is a signature of eukaryotic proteins . Here , we try to understand the origin of the difference in disorder between eukaryotic and prokaryotic proteins . We show that eukaryotic proteins contain more extended linker regions and that these linker regions are significantly more disordered . Further , we show , for the first time , that the difference in disorder originates from a systematic difference in amino acid frequencies between eukaryotic and prokaryotic proteins . Three amino acids contribute to the difference in disorder; serine and proline are more abundant in eukaryotic linkers , while isoleucine is less frequent . These shifts in frequencies are observed in all phyla , protein families , structural regions and type of protein but are most pronounced in disordered and linker regions . It is tempting to speculate that the increase in serine frequencies in eukaryotes is related to regulation by kinases , but direct evidence for this is lacking . Anyhow the widespread of the shifts in abundance indicates that the differences are ancient and caused be some yet not fully understood selective difference acting on eukaryotic and prokaryotic proteins .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"and",
"discussion"
] |
[
"chemical",
"compounds",
"aliphatic",
"amino",
"acids",
"organic",
"compounds",
"serine",
"isoleucine",
"amino",
"acids",
"bacteria",
"proteins",
"prokaryotic",
"cells",
"chemistry",
"cyclic",
"amino",
"acids",
"proline",
"biochemistry",
"eukaryota",
"organic",
"chemistry",
"cell",
"biology",
"proteomes",
"protein",
"domains",
"hydroxyl",
"amino",
"acids",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"organisms"
] |
2019
|
Why do eukaryotic proteins contain more intrinsically disordered regions?
|
The gene encoding a DNA/RNA binding protein FUS/TLS is frequently mutated in amyotrophic lateral sclerosis ( ALS ) . Mutations commonly affect its carboxy-terminal nuclear localization signal , resulting in varying deficiencies of FUS nuclear localization and abnormal cytoplasmic accumulation . Increasing evidence suggests deficiencies in FUS nuclear function may contribute to neuron degeneration . Here we report a novel FUS autoregulatory mechanism and its deficiency in ALS-associated mutants . Using FUS CLIP-seq , we identified significant FUS binding to a highly conserved region of exon 7 and the flanking introns of its own pre-mRNAs . We demonstrated that FUS is a repressor of exon 7 splicing and that the exon 7-skipped splice variant is subject to nonsense-mediated decay ( NMD ) . Overexpression of FUS led to the repression of exon 7 splicing and a reduction of endogenous FUS protein . Conversely , the repression of exon 7 was reduced by knockdown of FUS protein , and moreover , it was rescued by expression of EGFP-FUS . This dynamic regulation of alternative splicing describes a novel mechanism of FUS autoregulation . Given that ALS-associated FUS mutants are deficient in nuclear localization , we examined whether cells expressing these mutants would be deficient in repressing exon 7 splicing . We showed that FUS harbouring R521G , R522G or ΔExon15 mutation ( minor , moderate or severe cytoplasmic localization , respectively ) directly correlated with respectively increasing deficiencies in both exon 7 repression and autoregulation of its own protein levels . These data suggest that compromised FUS autoregulation can directly exacerbate the pathogenic accumulation of cytoplasmic FUS protein in ALS . We showed that exon 7 skipping can be induced by antisense oligonucleotides targeting its flanking splice sites , indicating the potential to alleviate abnormal cytoplasmic FUS accumulation in ALS . Taken together , FUS autoregulation by alternative splicing provides insight into a molecular mechanism by which FUS-regulated pre-mRNA processing can impact a significant number of targets important to neurodegeneration .
Amyotrophic lateral sclerosis ( ALS ) is a neuronal degenerative disorder caused by progressive loss of motor neurons in brain and spinal cord , leading to paralysis and death [1] . FUS is a frequently mutated gene in ALS ( combining familial and sporadic ALS ) , in addition to C9ORF72 , SOD1 and TDP-43 [1]–[3] . Most ALS-associated FUS mutations are within the nuclear localization signal ( NLS ) in the carboxyl terminus [2] , [4] , [5] , resulting in increased cytoplasmic FUS localization [6] , [7] . The abnormal cytoplasmic aggregation of FUS mutants in neuron and glial cells is a pathological hallmark of ALS and some cases of frontotemporal lobar degeneration ( FTLD ) [8]–[10] . It's noteworthy that there is a correlation between the observed cytoplasmic FUS accumulation and the age of ALS onset , with the more cytoplasmic FUS accumulation the earlier age of disease onset [8] , [11]–[13] . Several studies suggest that cytoplasmic accumulation of FUS mutant protein can lead to direct cytoplasmic cytotoxicity or may indirectly result in the loss of FUS function in the nucleus . Studies in yeast models demonstrated that expression of ALS-associated FUS mutants can lead to protein aggregation and cytotoxicity that recapitulate FUS proteinopathy [14] . Investigations in some Drosophila , C . elegans and rat models showed that expression of ALS-associated FUS mutants can lead to motor neuron dysfunction and neurodegeneration [15]–[17] . However , some Drosophila and zebrafish models support that the loss of FUS function can lead to behavioral and structural defects of motor neurons [18] , [19] . Exactly how the loss of FUS nuclear function and/or the gain of cytoplasmic cytotoxicity contribute to neurodegeneration at the molecular level is still unknown . FUS is predominantly a nuclear protein [20] , and binds both DNA and RNA [21] , [22] . FUS is involved in multiple steps of RNA metabolism including transcription , pre-mRNA splicing and mRNA transport for site specific translation [23]–[25] . The alteration of FUS-regulated RNA processing is a proposed key event in ALS pathogenesis , given that RNA binding proteins and splicing misregulation are linked to neurological diseases [8] , [26] , [27] . To understand the normal function of FUS in RNA processing , it is essential to identify FUS RNA targets . Recently a large number of FUS RNA targets in various cell lines and neural tissues were identified by CLIP-seq ( cross-linking and immunoprecipitation , followed by high-throughput sequencing ) , a method to purify protein-RNA complexes coupled with deep sequencing [28]–[32] . The challenge now is to begin to understand what the biological significance of FUS-regulated RNA processing is , and how these processes are altered in FUS mutants and may therefore contribute to ALS pathogenesis . Our CLIP-seq data in HeLa cells show that FUS binding is enriched in introns flanking cassette exons of pre-mRNAs . Among the identified FUS-binding cassette exons and their flanking introns , the most highly enriched target is exon 7 and flanking introns of FUS pre-mRNA itself . Here we demonstrate that FUS is a repressor of exon 7 and that the exon 7-skipped splice variants of FUS are subject to nonsense-mediated decay ( NMD ) . Overexpression of FUS leads to repression of exon 7 splicing and predictably a reduction of endogenous FUS protein levels . Conversely , knockdown of FUS protein reduces the repression of exon 7 . Moreover , the reduction of exon 7 repression can be rescued by the expression of EGFP-FUS . Taken together , these studies show that FUS dynamically autoregulates its own protein levels by directly modulating the alternative splicing of exon 7 . Furthermore , our data show that ALS-associated FUS mutants are deficient in nuclear localization , exon 7 repression and autoregulation of its own protein levels . We propose that the compromised FUS autoregulation in ALS forms a feed-forward loop , exacerbating the abnormal cytoplasmic FUS accumulation , and as such , provides a molecular mechanism that can potentially contribute to ALS pathogenesis .
To identify RNA targets of FUS , we performed FUS CLIP-seq in HeLa cells . Western blot and autoradiography showed successful immunoprecipitation of FUS protein and FUS-RNA complexes ( Figure 1A , 1B and 1C ) . Sequencing of FUS CLIP RNA yielded 1 , 879 , 212 non-redundant reads mapped to human genome GRCh37 , with the majority ( 1 , 305 , 507 reads ) to pre-mRNAs ( Figure S1 ) . Using the peak-finding algorithm CisGenome ( www . biostat . jhsph . edu/~hji/cisgenome/ ) [33] , we identified 1928 FUS CLIP clusters ( sites containing significantly enriched overlapping FUS CLIP tags ) corresponding to 1149 target genes ( Table S1 ) in HeLa cells . FUS RNA targets identified by our CLIP-seq were compared with those previously identified by other CLIP-seq [28]–[30] , [32] , PAR-CLIP [31] and RIP-chip [34] ( Figure S2 ) . HeLa and HEK293 cells [31] share 845 common target genes ( Figure S2 , Table S2 ) , accounting for 74% of all identified targets in HeLa cells . Gene Ontology ( GO ) biological process ( BP ) analysis of these 845 genes showed an enrichment of genes regulating gene expression and transcription . Analysis of various CLIP-seq datasets of mouse brains or neurons [28]–[30] , [32] identified 508 common genes , which are enriched for genes regulating cell adhesion , synaptic transmission , glutamate signaling pathways and nervous system development . 120 genes are common to all the datasets analyzed , and show an enrichment of genes regulating cell motion and protein dephosphorylation . Taken together , our analyses revealed cell-type specific and common RNA targets of FUS . Seventy five percent of our FUS CLIP clusters were located within introns ( Figure 1D ) , consistent with previous reports [28]–[32] . To address the function of FUS in alternative splicing , we analyzed the association between FUS CLIP clusters and known alternative splicing events . Using the UCSC Known AltEvent database as a reference , we scored a FUS CLIP cluster as associated with an alternative splicing event if the CLIP cluster overlapped the alternative splicing event itself or overlapped its immediate flanking introns and/or exons , as described previously [35] . Our analysis identified “cassette exon” as the top category of alternative splicing events associated with FUS CLIP clusters ( Figure S3 ) . FUS CLIP clusters are associated with 206 cassette exons in total ( Figure S3 ) . To identify FUS binding regions flanking cassette exons , we used 87 FUS-associated cassette exons that are flanked by constitutive exons ( Table S3 ) to generate a normalized complexity map as previously described [36] . We found that FUS binding was enriched in the flanking introns , particularly proximal to splice sites flanking the cassette exons ( Figure 1E ) . The peak at 5′ splice sites , within 100 nucleotides ( nt ) downstream of the cassette exons , showed the highest enrichment of FUS CLIP tags . The peak proximal to 3′ splice sites was about 150 nt upstream of the cassette exons , instead of immediately upstream of 3′ splice sites ( less than 50 nt ) , as previously described [31] . We also observed a peak at about 400 nt downstream of the cassette exons and a peak at about 300 nt downstream of the upstream constitutive exons . The locations of all these four peaks in our complexity map were also detected as statistically significant FUS binding sites in the complexity map from Lagier-Tourenne et al . [30] . Comparing our FUS complexity map with all the other reports [28]–[30] , it is consistent that in general FUS binding is enriched in the intronic regions 500 nt upstream or downstream of cassette exons or constitutive exons flanking cassette exons , although the exact nucleotide positions are not identical in different studies . FUS-RNA binding may be both position and sequence dependent . We next analyzed the sequences of FUS CLIP clusters associated with cassette exons and their flanking introns for possible de novo consensus RNA-binding motifs using the HOMER algorithm [37] . Analysis of the CLIP clusters within each individual peak on the complexity map ( Figure 1E ) did not identify any significant common consensus motifs ( data not shown ) . Individually , the highest FUS-binding peak at the 5′ splice sites downstream of cassette exons did show an enrichment of CAGGUU ( 2 . 6 fold , P = 0 . 001 ) ( Figure S4 ) ; however , this is expected as CAGGUU is very similar to the human 5′ splice site consensus sequences MAG|GURAGU ( M is A or C and R is A or G ) [38] . To assess whether genes encoding FUS-associated cassette exons can be clustered into functional groups , we analyzed the Gene Ontology ( GO ) biological process ( BP ) terms and KEGG pathways using DAVID Bioinformatics Resources 6 . 7 [39] . Our results showed that the most enriched GO BP terms were regulation of transcription , RNA metabolic process and neurogenesis ( Table S4 ) . The most enriched KEGG pathways were Wnt , adherens junction and Notch signaling pathways ( Table S5 ) . Out of the 87 cassette exons enriched with FUS CLIP clusters , the top candidate was exon 7 and flanking introns 6 and 7 of the pre-mRNAs of FUS itself ( Table S3 ) . This CLIP cluster was also in the top 10 of all 1928 FUS CLIP clusters identified , as ranked by fold enrichment . The number of FUS CLIP tags within exon 7 and its flanking introns were 8 . 1 fold higher than the control mouse IgG CLIP tags ( FDR = 0 . 043 ) , as determined by the peak finding algorithm CisGenome [33] ( Figure 2A ) . The region encompassing FUS intron6-exon7-intron7 is ∼3 kb and highly conserved in 38 vertebrate species ( Figure 2B ) . Human and mouse DNA sequences share 77% identity within this region , while the average similarity throughout other introns of FUS is 40% . Sub-regions with highly enriched CLIP tags ( over 100 overlapping CLIP tags in the center ) were used for de novo consensus RNA motif analysis . Analysis of all the CLIP tags within these selected regions using the Homer algorithm revealed that GU or GGU containing sequences are statistically enriched over background ( all pre-mRNA sequences ) ( Figure S5 ) . This is consistent with previous reports that GGUG , GUGGU , or GGU containing RNA sequences are potential FUS binding motifs [29] , [30] , [40] . To validate the interaction between FUS and its own pre-mRNA , we performed anti-FUS immunoprecipitation in HeLa cells , followed by RNA purification , reverse transcription ( RT ) and PCR using primers specific to the introns flanking exon 7 . Our results showed that indeed FUS interacted in vivo with its own pre-mRNAs at the region of exon 7 and its flanking introns ( Figure 2C ) , compared to the constitutive exon 5 that was not enriched in FUS immunoprecipitates . Of note , FUS-exon7 complex ( Figure 2C , lane 5 ) was immunoprecipitated even in the absence of UV crosslinking and under high salt wash conditions ( containing 750 mM NaCl ) , suggesting a strong association of FUS protein with exon 7 of its own pre-mRNAs . Skipping of FUS exon 7 results in an open reading frame shift and introduces a premature stop codon in exon 8 . The exon 7-skipped FUS transcripts ( NCBI RefSeq , NR_028388 . 2; Ensemble , ENST00000566605 ) are predicted to be subject to NMD . To detect the exon 7-skipped variant , we treated cells with cycloheximide ( CHX ) for 6 hours ( h ) , which inhibits translation and thereby NMD [41] . FUS splice variants were assessed by reverse transcription ( RT ) and PCR using primers in exon 6 and exon 8 . Indeed , we observed that exon 7-skipped splice variants were present in a variety of human and mouse cell lines including human cervical cancer cells HeLa , embryonic kidney cells HEK293 , neuroblastoma cells SH-SY5Y , and mouse motor neuron cells NSC-34 ( Figure 2D ) . Although the ratio of exon 7-skipped variants varies in different cell lines , they were all increased after CHX treatment , suggesting these variants undergo NMD . The PCR products of exon 7-skipped variants in HEK293 cells were confirmed by cloning and sequencing . At 6 h post CHX treatment , no significant changes were detected at the FUS protein level in all the cells tested ( Figure S6 ) . We assessed the splicing of FUS exon 7 in the context of the splicing reporter human β globulin minigene ( Figure 3A ) by RT-PCR [42] . In the pDUP-FUS-E7L ( Long ) construct , we cloned FUS exon 7 and 2 . 8 kb of flanking introns 6 and 7 to encompass the entire conserved region enriched with FUS CLIP tags . In the pDUP-FUS-E7S ( Short ) construct , we cloned FUS exon 7 with about 300 bp of each flanking intron to assess the effects of distal intronic regions without compromising exon 7 splice sites . The pDUP-FUS-E5 construct containing FUS exon 5 and its flanking introns was used as a control , since exon 5 is a constitutive exon with no enrichment of FUS binding . The level of exon 7 inclusion is around 50% in the context of pDUP-FUS-E7L reporter transfected into HEK293 cells ( Figure 3B , lane 1 ) , similar to the endogenous FUS transcript . Interestingly , exon 7 inclusion level was lower in the pDUP-FUS-E7S reporter ( Figure 3B , lane 3 ) , suggesting additional regulatory elements in intron 6 and intron 7 . This may explain why the entire region spanning intron6-exon7-intron7 is highly conserved . To assess whether the splicing of exon 7 is affected by FUS protein levels , a gain of function assay with the EGFP-FUS plasmids and a loss of function assay with FUS siRNA were performed in HEK293 cells . Our results showed that the repression of exon 7 was enhanced significantly in both pDUP-FUS-E7L ( from 48 . 3%±0 . 6% to 93 . 1%±2 . 7% , mean ± SD , n = 3 ) and pDUP-FUS-E7S reporters ( from 74 . 9%±1 . 3% to 91 . 0%±0 . 6% ) when EGFP-FUS was expressed ( Figure 3B , 3C ) . As a control , the splicing of exon 5 in the pDUP-FUS-E5 reporter was not affected by FUS overexpression . Conversely , the level of the exon 7-skipped products decreased strikingly in the pDUP-FUS-E7L reporter ( from 48 . 7%±4 . 5% to 3 . 7%±1 . 4% ) when endogenous FUS protein was reduced by siRNA ( Figures 3D , 3E ) . The exon 7-skipped splice variant in pDUP-FUS-E7S reporter was also decreased but to a lesser extent , from 64 . 8%±4 . 8% to 46 . 0%±4 . 1% , which suggests that more regulatory elements in the entire intron6-exon7-intron7 region are required to control exon 7 alternative splicing . To further test the dependence of exon 7 splicing on FUS protein levels , a rescue assay with the pDUP-FUS-E7L reporter was performed by knocking down endogenous FUS using siRNAs that target the 3′ UTR of FUS pre-mRNAs , followed by expressing EGFP-FUS ( Figure 4A , 4B ) . The level of exon 7-skipped splice variants was reduced from 67 . 7%±0 . 6% ( lane 1 ) to 9 . 5%±3 . 8% ( lane 3 ) at 48 h post siFUS treatment , and recovered to 82 . 0%±0 . 6% ( lane 9 ) after introduction of EGFP-FUS for 24 h ( Figure 4C ) . This assay strongly supports that FUS is a repressor of its own exon 7 . Moreover , it also demonstrated that EGFP-FUS is as competent as the endogenous FUS to repress exon 7 splicing . To determine whether FUS protein levels affect exon 7 splicing of endogenous FUS transcripts , semi-quantitative PCR using radiolabeled primers was performed to examine the FUS splice variants after siRNA knockdown . To prevent NMD of the exon 7-skipped FUS transcripts , we treated cells with cycloheximide ( CHX ) , which allowed visualization of increases of exon 7 skipping in endogenous transcripts ( Figure 5A , comparing bottom bands in lanes 4 and 5 with those in lanes 1 and 2 ) . It is important to note that the siRNA targets both FUS splice variants , but the difference in the reduction of each splice variant relative to its corresponding mock transfection control is quantifiable and informative [43] . The level of exon 7-skipped variants ( bottom band , lane 6 ) was reduced to 15% of the mock transfection ( bottom band , lane 4 ) , while the level of the exon 7-included variant ( top band , lane 6 ) was only reduced to 50% of the mock transfection control ( top band , lane 4 ) upon siRNA knockdown of FUS and CHX treatment ( Figure 5A ) . A lesser reduction in exon 7-included variants than in exon 7-skipped variants was also observed without CHX treatment ( lane 3 ) . This result is consistent with the splicing reporter minigene assay , indicating that FUS is a repressor of exon 7 and reduced FUS protein levels results in less exon 7 repression . Western blot analysis confirmed siRNA knockdown of endogenous FUS protein ( Figure 5B ) . Expression of splicing factors SF2 and hnRNPA1 were unaffected ( Figure 5B ) , suggesting changes in FUS splicing were unlikely the result of an indirect mechanism . Taken together , we have demonstrated that FUS is a key repressor of its own exon 7 in both splicing reporter assays and endogenous FUS splicing assays . We showed that FUS repressed its exon 7 splicing and that the resultant exon 7-skipped transcripts were degraded by NMD and cannot be translated to protein . This observation led us to hypothesize that FUS can autoregulate its own protein levels by regulating the alternative splicing of exon 7 . If our hypothesis is correct , we predict that exogenous expression of FUS will downregulate endogenous FUS protein levels by promoting exon 7 skipping and consequently NMD . Western blot analysis using FUS antibody detected both endogenous FUS and EGFP-FUS . The results showed that the endogenous FUS protein level was decreased by about 50% with transient expression of EGFP-FUS in HEK293 cells ( Figure 5C ) . The endogenous FUS mRNA level was measured by quantitative RT-PCR ( qRT-PCR ) using primers annealing to the 3′ UTR of endogenous FUS transcripts but not the coding sequence of the EGFP-FUS transcripts . There was a slight reduction of endogenous FUS mRNA levels in the EGFP-FUS expressing cells , but no statistical significance was detected ( Figure S7 ) , suggesting the observed reduction in FUS protein levels occurs mainly at the post transcriptional level . Our finding of FUS autoregulation is also consistent with the observation in a FUS transgenic mouse model that the endogenous mouse FUS protein was reduced following overexpression of human FUS [44] . The majority of ALS-associated FUS mutations occur within the region coding for the nuclear localization signal , resulting in cytoplasmic retention of FUS [5] and inferred loss of FUS function in the nucleus . We propose that FUS autoregulation is deficient in ALS mutants due to the alteration of their cellular localization , which results in compromised FUS-dependent splicing regulation . To test this hypothesis , we made EGFP-FUS constructs with the ALS-associated mutations R521G , R522G and ΔE15 ( deletion of last 12 amino acids in the C-terminus ) , which respectively correlates with minor , moderate , and severe cytoplasmic accumulation of FUS , as reported by others both in ALS patients and in cell culture systems [6] , [8] , [9] , [11] . As a control , an RNA binding incompetent mutant EGFP-FUS RRM 4F-L was made by mutating four phenylalanine ( F ) to leucine ( L ) ( F305L , F341L , F359L , and F368L ) in the RNA recognition motif ( RRM ) [45] . We tested the effects of these mutants on FUS cellular localization , exon 7 splicing and autoregulation in HEK293 cells . Consistent with previous reports [6] , [8] , [9] , [11] , we observed predominant nuclear localization of wildtype FUS , minor cytoplasmic and mainly nuclear localization of R521G , moderate cytoplasmic accumulation and aggregation of R522G , and severe cytoplasmic aggregation with much less nuclear localization of the FUS ΔE15 mutant , following transient transfection in HEK293 cells ( Figure 6A ) and mouse motor neuron cells NSC-34 ( Figure S8 ) . The RRM mutant , like the wildtype FUS protein , was predominantly localized in the nucleus . To test the function of the FUS mutants in exon 7 alternative splicing , the splicing reporter minigene pDUP-FUS-E7L together with either wildtype or mutant EGFP-FUS plasmids were transfected into HEK293 cells ( Figure 6B ) . Expression of wildtype EGFP-FUS protein resulted in repression of exon 7 as expected , with an increase of the exon7-skipped products from 40 . 4%±1 . 7% to 89 . 3%±3 . 0% ( mean ± SD , n = 3 ) . Expression of the ALS-associated mutants , compared to the wildtype FUS , resulted in significantly compromised repression of exon 7 ( P≤0 . 05 , n = 3 ) . The more cytoplasmic localization of the ALS mutants , the less exon 7 repression , with exon 7 skipping ratio of 87 . 6%±2 . 8% for R521G , 70 . 6%±2 . 4% for R522G and 33 . 3%±10 . 1% for ΔE15 mutant . Expression of the EGFP-FUS RRM 4F-L mutant only resulted in a mild reduction of exon 7 repression with a ratio of exon 7 skipping of 82 . 1%±4 . 3% , which was statistically different from the EGFP-FUS wildtype protein ( P≤0 . 05 , n = 3 ) . While the RRM 4F-L mutant was reported incompetent to bind RNA [45] , its expression did increase the nuclear concentration of FUS protein , which may result in more endogenous FUS binding the intron6-exon7-intron7 region in the reporter minigene and repression of the exon 7 splicing . This may explain the slight reduction of exon 7 repression when the RRM 4F-L mutant was expressed . Our data suggest that both the nuclear concentration of FUS protein and RNA binding are critical for the regulation of FUS exon 7 . Consistent with the splicing data , western blot analysis confirmed that increased exon 7 skipping led to a decrease in endogenous FUS protein levels by 55 . 5% , when EGFP-FUS wildtype protein was expressed ( Figure 6C ) . Conversely expression of various EGFP-FUS ALS mutants , which showed reduced nuclear localization and exon 7 skipping , resulted in less reduction of the endogenous FUS protein ( Figure 6C ) . Expression of the ΔE15 mutant that is predominantly localized in the cytoplasm only downregulated the endogenous FUS protein by 14 . 6% . This mild downregulation of FUS protein is consistent with little increase of exon 7 skipping observed in the splicing assay ( Figure 6B ) . We also confirmed in human neuroblastoma cells SH-SY5Y that ALS-associated FUS mutants were deficient in regulating exon 7 repression in the context of the pDUP splicing reporter minigene , and that the deficiency correlated with the extent of cytoplasmic localization of the mutants ( Figure S9 ) . We observed that wildtype endogenous FUS protein was co-localized with the cytoplasmic aggregates of FUS ΔE15 mutant in both HEK293 cells ( Figure 6D ) and mouse motor neuron cells NSC-34 ( Figure S8 ) , using an antibody which detects only the endogenous FUS protein but not the ΔE15 mutant by recognizing a C-terminal FUS epitope . These data suggest that the FUS mutants may sequester the wildtype FUS protein in the cytoplasm and further contribute to the aggregation . This is consistent with a recent report that GFP-FUS ALS mutant is co-localized with MYC-FUS wildtype protein in the cytoplasmic aggregates [46] . We report here the localization of the endogenous FUS protein in the cytoplasmic aggregates of ALS-associated FUS mutants . Taken together , our data suggest that the severity of FUS cytoplasmic accumulation correlates with the deficiency of exon 7 repression and autoregulation of FUS protein levels . The deficiency of ALS-associated FUS mutants in alternative splicing and autoregulation may exacerbate the cytoplasmic accumulation of ALS-associated FUS mutants . Our data showed FUS autoregulation is deficient in cells expressing ALS-associated FUS mutants . Moreover , this deficiency increases with the relative severity of the cytoplasmic accumulation of individual FUS mutants , which would be expected to exacerbate the rate of cytoplasmic accumulation and FUS proteinopathy . Use of splicing-modulating antisense oligonucleotides ( ASOs ) is a therapeutic strategy of great potential to treat diseases arising from splicing defects [47]–[49] . We rationalized that ASOs promoting FUS exon 7 skipping should mimic the repression of exon 7 by FUS and thereby have the potential to restore the deficient FUS autoregulation in patients with ALS-associated FUS mutations . We designed FUS-ASOs to target the junction of intron 6 and exon 7 . FUS-ASOs were synthesized using 2′-O-methyl-oligoribonucleotides with phosphorothioate linkages to increase ASO stability and then tested with the pDUP-FUS-E7L minigene in HEK293 cells . Our results showed that FUS-ASOs induced repression of exon 7 in a dose dependent manner ( Figure 7 ) . This suggests a possibility that deficient FUS autoregulation can be therapeutically restored to reduce or alleviate the extent of abnormal FUS cytoplasmic accumulation occurring in ALS patients with FUS mutations .
Here we report a novel autoregulatory mechanism of FUS by alternative splicing and NMD . The model shown in Figure 8A illustrates FUS autoregulation as a feedback loop to control the homeostasis of FUS protein levels . High levels of FUS protein lead to increased FUS binding to exon 7 and its flanking introns , promoting exon 7 skipping and NMD to reduce excessive FUS protein . Low levels of FUS protein would favor exon 7 inclusion , resulting in increased FUS protein production . Alternative splicing-mediated NMD and highly conserved intronic sequences represent an emerging common mechanism utilized by RNA binding proteins ( RBPs ) to maintain their homeostasis [43] , [50] , [51] . FUS now joins this increasing list of autoregulated RBPs , including PTB , hnRNP L , Nova and TDP-43 [43] , [50]–[52] . FUS regulates many aspects of gene expression including transcription , alternative splicing and RNA transportation [23]–[25] . Dynamic regulation and conserved targets suggest it is important to keep these functional activities of FUS in tight control , and that FUS likely has a co-factor role in coordinating them . For example , loss of FUS can cause genomic instability and developmental defects in mouse , Drosophila and Zebrafish [18] , [19] , [53] . Conversely , high levels of FUS are associated with cancer and ALS , and moreover , are known genetic determinants of these diseases . Overexpression of FUS is observed in liposarcoma and leukemia with FUS translocations [54] , [55] . Aberrant accumulation of FUS mutant protein is a characteristic pathology of FUS-associated ALS [8] , [9] . Depletion of FUS in the mouse nervous system affects the abundance or the splicing of about 1000 mRNAs [30] , suggesting that maintaining equilibrated FUS protein levels is critical for RNA processing . In ALS , another frequently mutated gene TDP-43 is also an RNA binding protein that autoregulates its own protein levels [52] , [56] . The direct mechanism of TDP-43 autoregulation is different from what we report here for FUS . TDP-43 binds to the 3′ UTR of its own pre-mRNA to trigger either NMD [56] or exosome-dependent degradation [52] . To our knowledge , we are the first to report a FUS autoregulatory mechanism through alternative splicing and NMD . Autoregulation of both FUS and TDP-43 by post transcriptional mechanisms suggests their functional activities are tightly controlled and that unbalancing of this regulation may underpin molecular mechanisms that promote neurodegeneration in ALS . Mice and rats expressing TDP-43 without the autoregulatory sequence developed more severe neurodegeneration than those expressing autoregulated wildtype or ALS-linked TDP-43 mutants , strongly suggesting deficient TDP-43 autoregulation contributes to neurodegeneration [57] . This is also likely the case for FUS autoregulation , which needs to be experimentally tested in rodent models . Interestingly , in both FUS-associated ALS and cancer , loss of heterozygosity of the FUS gene is never observed . This suggests , at least genetically , that while compromised FUS autoregulation contributes to the initiating or driving events resulting from FUS mutations in these diseases , the activity of the wild-type FUS allele is required or selected to maintain this pathological state . In this regard , it is important to understand how both mutant and wild-type FUS activities may contribute to the progression of ALS and cancer . Compromised FUS homeostasis by autoregulation is expected in cells harbouring ALS-associated mutations ( Figure 8B ) . The majority of ALS-associated mutations are located in the nuclear-localization signal ( NLS ) of FUS , resulting in both a cytoplasmic retention of FUS mutants and a reduction of FUS protein levels in the nucleus . The reduction of nuclear-localized FUS leads to the reduction of FUS exon 7 repression , which in turn likely induces production of more exon 7-included transcripts for translation , thereby driving elevated protein synthesis of FUS . In our model of FUS autoregulation , a deficiency of FUS in the nucleus would favour a feed-forward mechanism , and thereby actually exacerbate the abnormal cytoplasmic accumulation of FUS mutants . Our observation that endogenous wildtype FUS protein was co-localized with the cytoplasmic aggregates of FUS mutants in both HEK293 cells and NSC-34 motor neuron cells suggests that the FUS mutants may further sequester the wildtype FUS protein in the cytoplasm and form more aggregates . In ALS , FUS cytoplasmic accumulation is a progressive process and increases with disease duration [58] . The deficient FUS autoregulation may lead to long term detrimental effects , and could be part of the mechanism underlying age-dependent neurodegeneration and death of neurons with ALS-associated FUS mutants . Indeed , a genotype-phenotype relation between different FUS mutations and FUS cytoplasmic accumulation or the age of ALS onset is observed [8] , [11]–[13] . The stronger the NLS mutation ( severity of cytoplasmic retention ) , the earlier the age of ALS onset . The three FUS mutants we constructed , R521G , R522G and ΔE15 ( last 12 amino acids truncation ) represent minor , moderate and severe cytoplasmic accumulation , respectively . The reported mean age of disease onset is 43 for R521G , 28 . 5 for R522G and 18 for R495X ( last 32 amino acid truncation ) in the later generation [8] , [11] . Here , we demonstrated that in HEK293 cells and SH-SY5Y cells expressing these same R521G , R522G and ΔE15 FUS mutants , increased cytoplasmic localization of FUS directly correlated with increased deficiencies of exon 7 skipping and FUS autoregulation . We speculate that FUS exon 7-skipped splice variants are reduced in the tissues or cell lines derived from ALS patients with FUS mutations , which can be further experimentally tested . Regulated splicing of exon 7 is a good model to examine FUS-dependent alternative splicing in detail , since it is one of the most significant FUS CLIP clusters identified in our FUS CLIP-seq . Our finding is also consistent with the report that FUS binds to highly conserved introns of genes encoding RNA binding proteins [32] . FUS exon 7 and its flanking introns were also identified previously as RNA targets of FUS and of TDP-43 respectively from FUS CLIP-seq and TDP-43 CLIP-seq of mouse brains [30] , [56] , but the functional significance of this implicated region was not experimentally tested . Interestingly , the prime molecular target of two genetic determinants of ALS converges on the same highly conserved FUS alternative exon and its flanking introns . This makes a compelling argument that the processing of FUS pre-mRNA specifically , and by extension , the role of FUS in alternative exon splicing in general , is an important molecular determinant of ALS . Lagier-Tourenne et al . proposed that FUS binding to FUS intron6-exon7-intron7 may result in retention of intron 7 to make a shorter FUS transcript with an alternative 3′ UTR [30] . We noticed that both NCBI RefSeq database ( NR_028388 . 2 ) and Ensemble database ( ENST00000566605 ) annotated a FUS splice variant without exon 7 , which is predicted to undergo NMD . We experimentally validated that the exon 7-skipped variant of FUS was expressed in multiple human and mouse cell lines and that the steady state levels of the exon 7-skipped variant was increased after inhibition of NMD . Furthermore , we demonstrated that FUS is a repressor of its own exon 7 by splicing assays of both splicing reporter minigenes and endogenous FUS pre-mRNAs . FUS-regulated alternative splicing of cassette exons is not just limited to its own exon 7 . We found FUS CLIP clusters were significantly associated with alternative splicing events of cassette exons . Our normalized complexity map of 87 FUS-associated cassette exon events revealed that FUS CLIP clusters were enriched in the introns flanking cassette exons , proximal to upstream 3′ splice sites and downstream 5′ splice sites , with the highest peak overlapping the downstream 5′ splice sites . FUS binding proximal to 5′ splice sites suggests that FUS may be associated with the assembly of spliceosome at 5′ splice sites , consistent with previous reports that FUS , as well as the related family member TAF15 , are in the U1 snRNP ( small nuclear ribonucleoprotein ) complex [59] , [60] . FUS binding proximal to 3′ splice sites suggests it may also affect the spliceosome assembly at 3′ splice sites; this functional significance is yet to be determined [31] . Activation or repression of cassette exon splicing can be dependent on RNA binding positions , as suggested in Nova [36] and FOX2 [61] CLIP-seq data , which showed that Nova and FOX2 binding proximal to 3′ splice sites repressed cassette exons , while conversely binding proximal to 5′ splice sites promoted cassette exons . However , splicing factors such as hnRNP A1 and PSF binding to introns proximal to 5′ splice sites can also repress cassette exons [62] , [63]; which suggests that activation or repression of cassette exon splicing is likely more complex . We experimentally demonstrated in this paper that FUS is a repressor of its own exon 7 . However , it does not rule out the possibility that FUS may activate other cassette exons . The repression or activation of a given cassette exon by FUS in a tissue specific manner might be controlled by different signaling pathways and/or cell type specific splicing factors involved in the complex in a spatio-temporal manner . Sequence motif analysis of FUS CLIP clusters in our data set did not identify a significant common consensus motif ( data not shown ) . We found variable motifs throughout the four binding peaks in the complexity map ( data not shown ) , suggesting limited or context-dependent FUS binding specificity , consistent with previous reports [28] , [29] , [31] , [32] . Analysis of CLIP clusters within the region encompassed by the highest binding peak in the normalized complexity map ( 5′ splice site downstream of cassette exons ) did identify an enrichment of a CAGGUU motif , which is similar to the human 5′ splice site consensus sequence MAG|GURAGU [38] . While this might be expected , it is interesting to note that GGU is the most common FUS binding site consensus sequence . This was originally reported in a SELEX assay ( GGUG ) [40] , and subsequently by CLIP-seq analysis from Lagier-Tourenne et al . ( GUGGU ) [30] and Rogelj et al . ( GGU ) [29] . A detailed examination of consensus RNA motifs within FUS intron 6 and intron 7 did reveal that GU or GGU containing sequences were statistically enriched , which provides FUS intron6-exon7-intron7 as a model for further experimental determination of FUS binding sites . FUS CLIP-seq data and RNA-seq data revealed a wide range of pre-mRNAs as candidate targets of FUS [28]–[32] , [64] . However , the biological significance of FUS-regulated RNA processing is only now being examined . Here we demonstrated that FUS-regulated cassette exon splicing of its own pre-mRNA leads to NMD , suggesting that FUS-regulated alternative splicing may be a common post transcriptional mechanism for the regulation of gene expression . This function of FUS in regulating cassette exon splicing is likely conserved in different tissues and species , since FUS-associated cassette exons were also observed in human and mouse neural tissues [28]–[30] . Our evidence demonstrating FUS is a splicing repressor of its exon 7 strongly supports the hypothesis that alternative splicing of many other neuronal- and disease-associated genes containing FUS-targeted cassette exons may also be regulated by FUS . In conclusion , our study uncovers an autoregulatory mechanism of FUS expression through alternative splicing and NMD , and demonstrates that its function in splicing regulation is deficient in ALS-associated FUS mutants . This study addresses a biological significance of FUS-regulated alternative splicing , and its potential relevance to ALS pathogenesis . Furthermore , our findings have important implications for the development of new therapeutic approaches to target alternative splicing in treating ALS . Taking advantage of our findings , splicing-modulating antisense oligonucleotides can be developed to induce exon 7 skipping and produce the FUS splice variants undergoing NMD . This may be a promising strategy to reduce the abnormal FUS cytoplasmic accumulation in ALS . Moreover , FUS autoregulation by alternative splicing provides insight into a molecular mechanism by which FUS-regulated pre-mRNA processing can impact a significant number of targets important to neurodegeneration .
Human cervical cancer cell line HeLa ( ATCC , CCL-2 ) and human embryonic kidney cell line HEK293 ( ATCC , CRL-1573 ) were grown in DMEM with 10% bovine growth serum ( HyClone , SH30541 . 03 ) . Human neuroblastoma cell line SH-SY5Y ( gift from Dr . Louise Simard ) was cultured in MEM-F12 medium ( 1∶1 ) supplemented with 10% fetal bovine serum ( GIBCO , 12483 ) . Mouse motor neuron cell line NSC-34 ( gift from Dr . Louise Simard ) was cultured in DMEM with 10% fetal bovine serum . CLIP-seq was performed as described [65] . Briefly , HeLa cells on 15 cm dishes were UV crosslinked in vivo at 400 mJ/cm2 using Stratalinker ( Stratagene 1800 ) . Cell lysates were prepared in lysis buffer [65] , followed with partial RNase A ( Sigma , R6513 ) digestion at different final concentrations ranging from 0 . 001 µg/ml to 0 . 1 µg/ml . FUS-RNA complexes were immunoprecipitated using mouse monoclonal anti-FUS antibody 10F7 pre-bound with protein G agarose beads ( Pierce , 22851 ) . 10F7 is a mouse monoclonal antibody previously developed in the Hicks laboratory by immunizing BALB/c mice with GST ( glutathione-s-transferase ) -FUS fusion protein . Various clones were screened using western blotting , immunocytochemistry , and flow cytometry , and the 10F7 antibody performed best for all three methods . This antibody recognizes amino acids 34–51 of FUS . Immunoprecipitation using mouse IgG ( Sigma , I5381 ) prebound with protein G agarose beads was performed in parallel as a control . While FUS-RNA complexes were still bound to beads , the 5′ end of CLIP RNA was radiolabeled with [γ-32P] ATP , and the 3′ end of CLIP RNA was ligated to a 3′ RNA linker . The radiolabeled FUS-RNA complex was separated onto a 10% ( w/v ) Bis-Tris gel ( Novex NuPAGE ) , transferred to a nitrocellulose membrane ( Bio-Rad , 162-0115 ) , and exposed to X-ray film ( Amersham Hyperfilm MP ) for autoradiography . The appropriate shifted FUS-RNA bands were cut out of the nitrocellulose membrane and subject to protein digestion using proteinase K ( Roche , 3115879001 ) . RNA was recovered using phenol chloroform extraction and sodium acetate , ethanol-isopropanol ( 1∶1 ) precipitation . The recovered RNA was further ligated to a 5′ RNA linker , and subject to DNase ( Promega , M6101 ) digestion and RNA recovery again . The final RNA product was reverse transcribed to cDNA using linker specific primers and subject to deep sequencing using the Illumina Genome Analyzer II ( single end , 72 bp ) at the Center of Applied Genomics in Toronto ( TCAG ) . Sequences of unique CLIP tags were mapped to the human genome ( GRCh37 ) by BlastN [66] after trimming the CLIP linker sequences and removing duplicate CLIP tags . A peak finder algorithm CisGenome ( www . biostat . jhsph . edu/~hji/cisgenome/ ) [33] was used to define CLIP clusters with significant enrichment ( FDR≤0 . 05; FUS CLIP vs control mouse IgG CLIP ) . The RNA targets from our FUS CLIP-seq data in HeLa were compared to the RNA targets previously identified by other FUS CLIP-seq , PAR-CLIP and RIP-ChIP in different tissues and cells [28]–[32] , [34] . If the gene list was not reported in the paper , the raw data of deposited fastq files [28]–[30] were retrieved from ENA ( The European Nucleotide Archive , http://www . ebi . ac . uk/ena/ ) and uploaded to Galaxy ( http://main . g2 . bx . psu . edu/ ) [67] for sequence analysis . The CLIP-seq reads were mapped to mouse genome ( mm9 ) using Bowtie ( version 1 . 1 . 2 ) [68] built in Galaxy , with parameters reported in the corresponding paper . The same peak finder algorithm CisGenome was used to identify the CLIP clusters in all the datasets . Mouse genes were converted to the HGNC-approved human gene symbols ( http://www . genenames . org/ ) to facilitate the comparison of different datasets . To identify the overlapping and the non-overlapping genes between different datasets , lists of the genes from the datasets were loaded into Venn , a web-based Venn diagram program , or Venn Diagram Plotter , a PC-based Venn diagram program ( http://omics . pnl . gov/ ) . To identify our FUS CLIP clusters overlapping alternative splicing events , the genomic coordinates of CLIP clusters were searched against the UCSC Known AltEvent database ( hg 19 ) [69] as described previously [35] , [70] . Out of all 206 FUS-associated cassette exons , 87 cassette exons which are flanked by constitutive exons were used to generate a normalized complexity map as previously described [36] . FUS CLIP tags within 500 nucleotides upstream and/or downstream of these cassette exons and flanking exons were mapped . Control was an average of 100 sets of normalized complexity of 87 constitutive exons randomly selected from genes expressed in HeLa cells as determined by RNA-seq [71] . Analysis of de novo consensus RNA motif enrichment was performed using findMotifsGenome perl script of the Homer software [37] with parameters of 5 or 6 bases for motif length , 42 for target size , and –RNA option . Gene Ontology ( GO ) analysis was performed using DAVID Bioinformatics Resources 6 . 7 ( http://david . abcc . ncifcrf . gov/ ) [39] or Enrichr ( http://amp . pharm . mssm . edu/Enrichr/index . html ) [72] . Immunoprecipitation of FUS protein was performed using mouse monoclonal anti-FUS antibody ( 10F7 ) . FUS-bound RNA was recovered using phenol-chloroform extraction and sodium acetate , ethanol–isopropanol ( 1∶1 ) precipitation , as described previously [65] . DNA was removed by DNase treatment ( Ambion , AM1906 ) . The recovered RNA was reverse transcribed to cDNA using Superscript III ( Invitrogen ) and amplified using Phusion Hot Start Polymerase ( NEB ) . Primers used for amplifying the region flanking FUS exon 7: 5′-ACAACCTTTTGTAGCCGTTGGAAG-3′ ( forward ) , 5′-CTTTCTGGAGGTGGTTCTGGACAC-3′ ( reverse ) . Primers used for amplifying the region flanking FUS exon 5: 5′-TCCCTAGTTACGGTAGCAGTTCTC-3′ ( forward ) , 5′-GCTGCAGACAAAGCTGAAGACATC-3′ ( reverse ) . PCR products were resolved on a 2% agarose gel and visualized by ethidium bromide staining . Cycloheximide ( CHX; Sigma ) was added to cell culture medium at the final concentration of 100 µg/ml to inhibit translation and thereby nonsense mediated decay ( NMD ) . At 6 h post CHX treatment , cytoplasmic RNA was extracted using RNeasy kit ( QIAGEN ) as per manufacture's recommendation . Reverse transcription ( RT ) was performed using Superscript III reverse transcriptase ( Invitrogen ) . Radiolabeled PCR was performed to amplify FUS exon 7 splice using Phusion Hot Start DNA Polymerase ( NEB ) . Primers were designed to anneal to exon 6 and exon 8 . Primers: 5′-AGTGGTGGCTATGAACCCAGAGGT-3′ ( forward ) , 5′-AGTCATGACGTGATCCTTGGTCCC-3′ ( reverse ) . The reverse primer was labeled with [γ-32P] ATP using T4 PNK ( NEB ) . PCR products were resolved on a 6% polyacrylamide/8M urea denaturing gel . The gel was dried , exposed to a phosphorimager plate ( Kodak ) . The images of radioactivity signals were captured by a phosphorimager ( Bio-Rad , Personal FX ) . The density of the radioactive bands was quantified using the Image J program v1 . 44p ( NIH , Bethesda , MD , USA , http://rsbweb . nih . gov/ij/ ) . FUS cDNA was amplified from human fetal liver pAct2 cDNA library ( Clontech ) . EGFP-FUS expression construct was made by subcloning the open reading frame of FUS cDNA ( RefSeq: NM_004960 . 3 ) into the BglII and KpnI sites of pEGFP-C1 plasmid ( Clone Tech ) . Mutagenesis of ALS-associated mutations ( R521G , R522G ) and deletion ( ΔE15 ) were performed using Quickchange Lighting Site-directed Mutagenesis Kit as per manufacturer's recommendation ( Stratagene ) . The EGFP-FUS RNA recognition motif ( RRM ) mutant 4F-L ( F305L , F341L , F359L , and F368L ) was generated using the QuikChange Lightning Multi-Site-Directed Mutagenesis kit as per manufacturer's recommendation ( Stratagene ) . EGFP-FUS and mutant constructs were verified by DNA sequencing . EGFP or EGFP-FUS ( wildtype or mutant ) plasmids were transiently transfected into HEK293 cells , or NSC-34 cells using Lipofectamine 2000 ( Invitrogen ) reagent as per manufacturer's recommendation . Cells were fixed with 4% paraformaldehyde . To detect the endogenous FUS protein , cells were incubated with the primary antibody rabbit anti-FUS ( Bethyl Laboratories , BL1355 ) and the secondary antibody Alexa Fluor 568 or Alexa Fluor 488 donkey anti-rabbit antibody ( Invitrogen ) . Nuclei were counter stained with DAPI or NucRed Dead 647 dye ( Invitrogen ) . The localization of EGFP-FUS , mutants , and endogenous FUS protein was imaged using an Olympus FV500 confocal microscope and analyzed with Fluoview software version 4 . 3 ( Olympus ) . Line sequential scanning was applied to avoid the potential bleed-through of fluorescence . Splicing reporter minigene pDUP-FUS constructs: FUS exon 7 and its flanking introns were amplified using Phusion Hot Start Flex Polymerase ( NEB ) from human genomic DNA extracted from HEK293 cells . The sequence of interest was subcloned between the ApaI and BglII sites of the splicing reporter minigene pDUP175 plasmid [42] . Three reporters were made . The pDUP-FUS-E7L ( Long ) construct contains FUS exon 7 , and 1453 bp upstream and 1355 bp downstream of the flanking introns . The pDUP-FUS-E7S ( Short ) construct contains FUS exon 7 , 292 bp upstream and 321 bp downstream of the flanking introns . The pDUP-FUS-E5 construct is a control construct containing the sequence of FUS exon 5 and its flanking introns . Splicing reporter minigene assay with FUS overexpression: 0 . 5 µg of pDUP reporter and 2 µg of EGFP-FUS plasmid or EGFP-FUS mutants were transfected into HEK293 cells using Lipofectamine 2000 ( Invitrogen ) reagent as per manufacturer's recommendation . At 48 h post transfection , cytoplasmic RNA was purified using RNeasy kit ( QIAGEN ) . RT-PCR was performed to assess the splicing of FUS exon 7 using Superscript III reverse transcriptase ( Invitrogen ) and Phusion Hot Start DNA Polymerase ( NEB ) . Primer sequences: 5′-CTCAAACAGACACCATGCATGG-3′ ( forward ) and 5′-CAAAGGACTCAAAGAACCTCTG-3′ ( reverse ) . PCR products were resolved on a 3% agarose gel with ethidium bromide staining , and imaged using Fusion FX imager ( Vilber Lourmat ) . The intensity of PCR bands was quantified using ImageJ software as described above . Splicing reporter minigene assay with siRNA knockdown of FUS: 0 . 5 µg of the pDUP-FUS-E7L reporter and 20 nM ( final concentration ) of FUS siRNA ( Dharmacon ON-TARGETplus SMARTpool ) were transfected into HEK293 cells using Lipofectamine 2000 reagent ( Invitrogen ) as per manufacturer's recommendation . At 48 h post transfection , cytoplasmic RNA was purified; and splice variants were analyzed by RT-PCR and gel electrophoresis , as described above . Rescue assay: Endogenous FUS protein was knocked down by transfecting HEK293 cells with a custom designed siRNA targeting 3′ UTR of human FUS ( siFUS ) . siFUS sequence: 5′-UAUAGUUACAAUUACAUAGUCCGACAC-3′ ( IDT , DsiRNA ) . The siRNA was transfected using Lipofectamine 2000 ( Invitrogen ) as per manufacturer's recommendation . At 48 h post transfection of siRNA , cells were retransfected with pDUP-FUS-E7L plasmid alone or together with EGFP or EGFP-FUS plasmid to rescue the FUS protein levels . At 24 h post re-transfection , cytoplasmic RNA was isolated for analysis of splice variants by RT-PCR , as described above . Rabbit anti-FUS antibody ( Bethyl Laboratories , BL1355 ) or mouse anti-FUS antibody ( 10F7 ) , rabbit anti-Actin antibody ( Sigma , A2066 ) , mouse anti-SF2 ( Santa Cruz , SC73026 ) and mouse anti-hnRNP A1 ( ImmuQuest , IQ205 ) were used for western blot analysis . Western blot was developed using ECL prime reagent ( Amersham ) and imaged with ChemiDoc MP imaging system ( Bio-Rad ) . The protein band intensity was quantified using the ImageJ program v1 . 44p ( NIH , Bethesda , MD , USA , http://rsbweb . nih . gov/ij/ ) . Actin was used to normalize the loading of protein amount . The endogenous human FUS transcripts were measured by quantitative RT-PCR with qPCR SYBR Green mix ( Fermentas ) , using a real-time PCR system ( Bio-Rad , CFX96 ) . Primers were designed to anneal to the 3′ UTR region of the endogenous FUS transcript . Primer sequences: 5′-CCAATTCCTGATCACCCAAGGGTTT-3′ ( forward ) , 5′-TGGGCAGGGTAATCTGAACAGGAA-3′ ( reverse ) . 2′-O-methyl-oligoribonucleotides with phosphorothioate linkages were synthesized and purified by Trilink Biotechnologies , Inc . ( San Diego , CA ) . FUS-ASOs target the splice junction spanning intron 6 and exon 7 . FUS-ASO sequences: 5′-GUCACUUCCGCUGGAGAAGA-3′ . Control ASOs ( Ctrl-ASOs ) are ASOs targeting the SRA gene [73] . ASOs alone ( final concentration 2 nM , 10 nM and 50 nM ) or ASOs together with pDUP-FUS-E7L reporters ( 0 . 5 µg ) was transfected into HEK293 cells using Lipofectamine 2000 reagent ( Invitrogen ) as per manufacturer's recommendation . The splicing of exon 7 in the reporter was assessed by RT-PCR , as described above .
|
FUS/TLS is a frequently mutated gene in amyotrophic lateral sclerosis ( ALS ) . ALS , also known as Lou Gehrig's disease , is characterized by a progressive degeneration of motor neurons . The abnormal cytoplasmic accumulation of mutant FUS protein is a characteristic pathology of ALS; however , recent evidence increasingly suggests deficiencies in FUS nuclear function may also contribute to neurodegeneration in ALS . Here we report a novel autoregulatory mechanism of FUS by alternative splicing and nonsense mediated decay ( NMD ) . We show FUS binds to exon 7 and flanking introns of its own pre-mRNAs . This results in exon skipping , inducing a reading frame shift and subsequent degradation of the splice variants . As such , this mechanism provides a feedback loop that controls the homeostasis of FUS protein levels . This balance is disrupted in ALS-associated FUS mutants , which are deficient in nuclear localization and FUS-dependent alternative splicing . As a result , the abnormal accumulation of mutant FUS protein in ALS neurons goes unchecked and uncontrolled . Our study provides novel insight into the molecular mechanism by which FUS regulates gene expression and new understanding of the role of FUS in disease at the molecular level . This may lead to new potential therapeutic targets for the treatment of ALS .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
ALS-Associated FUS Mutations Result in Compromised FUS Alternative Splicing and Autoregulation
|
Bats are a highly successful , globally dispersed order of mammals that occupy a wide array of ecological niches . They are also intensely parasitized and implicated in multiple viral , bacterial and parasitic zoonoses . Trypanosomes are thought to be especially abundant and diverse in bats . In this study , we used 18S ribosomal RNA metabarcoding to probe bat trypanosome diversity in unprecedented detail . Total DNA was extracted from the blood of 90 bat individuals ( 17 species ) captured along Atlantic Forest fragments of Espírito Santo state , southeast Brazil . 18S ribosomal RNA was amplified by standard and/or nested PCR , then deep sequenced to recover and identify Operational Taxonomic Units ( OTUs ) for phylogenetic analysis . Blood samples from 34 bat individuals ( 13 species ) tested positive for infection by 18S rRNA amplification . Amplicon sequences clustered to 14 OTUs , of which five were identified as Trypanosoma cruzi I , T . cruzi III/V , Trypanosoma cruzi marinkellei , Trypanosoma rangeli , and Trypanosoma dionisii , and seven were identified as novel genotypes monophyletic to basal T . cruzi clade types of the New World . Another OTU was identified as a trypanosome like those found in reptiles . Surprisingly , the remaining OTU was identified as Bodo saltans–closest non-parasitic relative of the trypanosomatid order . While three blood samples featured just one OTU ( T . dionisii ) , all others resolved as mixed infections of up to eight OTUs . This study demonstrates the utility of next-generation barcoding methods to screen parasite diversity in mammalian reservoir hosts . We exposed high rates of local bat parasitism by multiple trypanosome species , some known to cause fatal human disease , others non-pathogenic , novel or yet little understood . Our results highlight bats as a long-standing nexus among host-parasite interactions of multiple niches , sustained in part by opportunistic and incidental infections of consequence to evolutionary theory as much as to public health .
Trypanosoma cruzi is the etiological agent of Chagas disease , a complex zoonosis that continues to take dozens of human lives each day [1] . Alongside its close relative Trypanosoma cruzi marinkellei in the Schizotrypanum subgenus , this important protozoan flagellate belongs to a broader , inter-continental group ( the “T . cruzi clade” ) of ancient endoparasites found to infect the mammalian fauna far and wide [2–3] . Infections have been reported in primates of Africa [4] , marsupials of Australia [5] and a multitude of terrestrial mammals across the Americas [6] , but most of this striking spread in host diversity tallies to few taxa within the clade ( above all to T . cruzi sensu stricto , i . e . , T . cruzi , and to T . rangeli ) . The majority of T . cruzi clade diversity is found in bats . Chiroptera are known to carry both generalists such as T . cruzi and T . rangeli as well as multiple bat-restricted species—some abundant ( e . g . , T . c . marinkellei , T . dionisii and T . erneyi ) , others rare ( e . g . T . livingstonei and T . wauwau ) [3 , 7–8] . Chiropteran immunity is unique with respect to other mammalian genera , coincident perhaps with physiological adaptations to flying [9] . Several features of bat immunity may predispose bats to long-term asymptomatic infections [10] with viruses [11–12] , bacteria [13–14] , fungi [15–16] , protozoa [17–18] and helminths [19–20] , several of which cause disease in humans and animals [21] . Given the diversity of bat-infecting T . cruzi-clade trypanosomes throughout the New and Old Worlds , many now accredit the Chiroptera with a fundamental role in the evolution of this parasite group [22] . In fact , the most parsimonious explanation to date for the origin and past expansion of the T . cruzi clade suggests a common ancestral lineage of bat-restricted trypanosomes that diversified into several independent lineages that on rare occasion switched into other terrestrial mammal hosts [17] . Bats’ recurrent interaction with other mammals and their various ectoparasites are thought to have afforded enough opportunity for at least five such switching or “seeding” events , likely since the early Eocene ( 54 to 48 million years ago ) [7] . Many trypanosomes from bats are morphologically indistinguishable , often described simply as “T . cruzi-like” in the past [23] . As mixed species/genotype infections are probably common but overlooked or mistaken , molecular barcoding presents expedient recourse in resolving intricate trypanosomatid taxonomy and ecology . Metabarcoding couples classic molecular barcoding with next generation sequencing techniques [24–25] to generate thousands of sequence reads from a single sample [26–27] . These reads correspond to the diversity and abundance of organisms infecting the host individual [28–30] . In this study , we applied next-generation metabarcoding methods to the most bat-diverse ( per area ) biome of Brazil [31] . We focused on a degraded section of Atlantic Forest in Espírito Santo ( ES ) state where terrestrial mammals appear reduced in abundance as well as in T . cruzi infection . A fatal case of human T . cruzi ( I-IV ) and T . dionisii coinfection [32] immediately predated the bat trypanosome survey by 18S ribosomal RNA deep sequencing in this region .
The sampling procedures reported herein were authorized by the Brazilian Institute of the Environment and Renewable Natural Resources ( IBAMA ) under license no . 19037–1 ( 23-05-2009 ) . Euthanasia and blood collection met guidelines set by the Federal Council of Veterinary Medicine , Resolution 1000 ( 11-05-2012 ) , in accordance to Federal Law 11 . 794/2008 . All procedures followed protocols approved by the Oswaldo Cruz Foundation ( Fiocruz ) Ethics Committee for Animal Research ( L0015-07 ) . Bat captures were carried out in two periods of 2015: June ( dry season ) and November ( rainy season ) . Mist nets were opened upon sunset for four hours on two consecutive nights at each study location . A total of 108 bats were captured using ten mist nets ( 3 x 9 m , 35 mm mesh ) placed along forest edges near banana and coffee crops at three different rural locations in Guarapari municipality , ES state , southeast Brazil: Rio da Prata ( 350 m a . s . l . ) , where a fatal case of Chagas disease occurred in 2012; Buenos Aires ( 250 m a . s . l . ) , where reports of triatomine invasion have increased in recent years; and Amarelos ( at sea level ) , where triatomines have not been reported from the domestic zone ( based on records by the Zoonosis Control Center , Guarapari municipality , ES ) ( S1 Fig ) . Taxonomic identification by morphology followed [33] and a maximum of ten individuals per species ( per site ) were kept for further sampling , as specified by law . Once anesthetized with acepromazine ( 2% ) in 9:1 ketamine hydrochloride ( 10% ) , these individuals were cleared of fur in the pectoral region ( by scalpel ) and sterilized with antiseptic soap and iodinated ethanol ( 70% ) for blood withdrawal by cardiac puncture . Within the safety area of a flame , 300 μl blood was collected into sterile 1 . 5 ml vials and stabilized in two parts ( i . e . , 600 μl ) 6 M Guanidine-HCl , 0 . 2 M EDTA solution for storage at -20°C . All bats used in these analyses received a collection number with the initials of the collector ( RM ) and were prepared for fluid preservation . This material will be subsequently deposited at the mammal collection of Museu Nacional , Federal University of Rio de Janeiro , Rio de Janeiro , Brazil . DNA was purified from 90 guanidine-EDTA blood lysates in DNeasy mini spin columns ( Qiagen ) , with each of nine extraction rounds including one negative control . Purified DNA samples were then PCR-amplified with primers 5’-TGGGATAACAAAGGAGCA-3’ ( forward ) and 5’-CTGAGACTGTAACCTCAAAGC-3’ ( reverse ) for 30 cycles of 94°C ( 30 s ) , 55°C ( 60 s ) and 72°C ( 90 s ) to target a trypanosome-specific , ~556 bp region of the 18S rRNA gene as established in [5] . For a subset of samples , a wider , ~927 bp region ( encompassing the ~556 bp above ) was first targeted with external primers 5’-CAGAAACGAAACACGGGAG-3’ ( forward ) and 5’-CCTACTGGGCAGCTTGGA-3’ ( reverse ) at equivalent cycling conditions to form a nested ( two-round ) PCR amplification procedure following [34] . Sterile water ( 2x ) and sample-free eluate from prior DNA purification ( 1x ) were used to provide three negative controls per 20-sample PCR reaction . Amplicons were single-end barcoded [35] , purified by agarose gel electrophoresis ( PureLink Quick Gel Extraction Kit , Invitrogen ) , quantified by fluorometric assay ( Qubit 2 . 0 , Thermo Fisher Scientific ) and pooled to equimolar concentration for multiplexed , paired-end ( 2 x 300 bp ) sequencing on the Illumina MiSeq platform ( Reagent Kit v2 ) . Amplicon sequences were filtered to retain only full-length reads of ≥ 99 . 9% base call accuracy by windowed trimming in Sickle [36] , verified for quality in FastQC [37] and mapped against a Trypanosoma spp . reference collection from SILVA v119 [38] using Bowtie 2 [39] . Operational Taxonomic Unit ( OTU ) construction proceeded by UPARSE algorithm in USEARCH [40] and BLAST-based taxonomic assignment in the QIIME environment [41] , with run parameters established during prior in silico testing on trypanosomatid 18S rRNA sequences from NCBI . Samples were clustered to OTUs de novo at 98% sequence similarity and assigned to extant species with a confidence threshold of 80% . Unassigned clusters were considered valid OTUs only if present at > 300 reads in any single sample and present at > 600 reads across all samples of the dataset . Following OTU establishment , sequence read pairs from one representative per OTU were merged and aligned in Clustal W ( with manual refinement of misplaced reads ) . Phylogenies were inferred in Mega 6 [42] by maximum likelihood ( ML ) tree construction under Kimura’s two-parameter model of nucleotide substitution with gamma-distributed variation among sites ( K2 + G ) . One thousand bootstrap replicates were run to establish nodal support . The 50 18S rRNA reference sequences applied in phylogenetic analyses are listed with accession numbers in S1 Table . All sequences have been deposited in the NCBI Sequence Read Archive ( SRA ) under accession numbers SRR5451077-SRR5451120 .
Of the 108 bats captured at Amarelos , Buenos Aires and Rio da Prata study sites , 105 individuals represent 16 species in the Phyllostomidae family , and three individuals represent one species ( Myotis nigricans ) in the Vespertilionidae family . Species and their abundances are listed in Table 1 . Standard and/or nested PCR amplified 18S rRNA gene fragments from 34 of 90 ( 38% ) bat blood samples . The 34 positive samples derived from 13 bat species ( of 17 species analysed ) and comprised 14 distinct kinetoplastid OTUs . Five OTUs were assigned to T . cruzi I ( OTU 3 ) , T . cruzi III/IV ( OTU 5 ) , T . c . marinkellei ( OTU 6 ) , T . rangeli lineage D ( OTU 10 ) and T . dionisii ( OTU 2 ) . A further seven OTUs did not assign to any known species of the T . cruzi clade . Phylogenetic analyses placed these seven OTUs ( 1 , 7 , 8 , 11 , 12 , 13 and 14 ) within a monophyletic group that includes trypanosome species from bats of the New World . Finally , two OTUs showed greater homology outside of the T . cruzi clade—OTU 4 , similar to a trypanosomatid species found in reptiles , and OTU 9 , nearly identical to the eubodonid Bodo saltans ( Figs 1 and 2 , S1 Table ) . Most trypanosome-infected bats presented mixed infections by two to eight OTUs . Only three positive blood samples ( from D . rotundus , G . soricina and R . pumilio ) contained a single OTU ( T . dionisii; OTU 2 ) . The bat species A . lituratus , C . perspicillata , D . rotundus and P . recifinus presented greatest trypanosome diversity , with seven to eight OTUs per species ( Fig 3 ) . Across the three study sites , trypanosomatid diversity and abundance broadly reflected bat capture success rather than any feature of the capture environment ( Table 1 and Fig 4 ) . Nested PCR detected between one and six more OTUs than standard PCR in eight of ten samples subjected to both procedures , showing less sensitivity only in samples RM 847 and RM 2009—one and two less OTUs amplified , respectively ( Fig 3 ) .
In this study , we exposed unforeseen bat trypanosome 18S rRNA diversity from standard capture effort in Atlantic Forest fragments of Guarapari municipality , ES , southeast Brazil . Our metabarcoding approach identified a preponderance of coinfection , involving several human-pathogenic and bat-associated types of the T . cruzi clade , as well as a swathe of yet undescribed diversity closer to its base . Furthermore , we identified sequences from two divergent kinetoplastid taxa—one similar to trypanosomatid isolates from reptiles , another matching the non-parasitic B . saltans . Unprecedented as they may be as complex co-infections , the diversity of individual kinetoplastids we report is not unexpected . Every recent trypanosome survey of bats has revealed novel parasite genotypes , host- and/or geographic range [8 , 43–50] , with particular surges in discovery following intensified sampling ( e . g . , transcontinental archival analysis ) [8] or innovative approach ( e . g . , coalescent species delimitation ) [43] . The 18S rRNA deep sequencing in bats here identifies further diversity around the most basal T . cruzi clade trypanosomes of the New World , with seven independent and novel taxonomic units forming sister groups to T . wauwau and Neobat species found in mormoopid and phyllostomid bats [8] . This expansion of a group related more closely to trypanosomatids detected in Australian marsupials than to those known from other neotropical mammals’ points to the Chiroptera as an ancient , perhaps original host order of the T . cruzi clade . Our data reinforce the bat host range of T . cruzi-clade trypanosomes across frugivorous , nectarivorous , carnivorous , generalist and hematophagous phyllostomid genera ( Anoura , Artibeus , Carollia , Desmodus , Glossophaga , Platyrrhinus , Phyllostomus , Rhinophylla , Sturnira , Trachops ) and into the ( primarily insectivorous ) Vespertilionidae . Our study provides strong , if circumstantial , evidence for the role of bats as T . cruzi reservoirs in ES state . Trypanosoma cruzi I and III/V found in bats of this study correspond to Discrete Typing Units ( DTUs ) associated with a recent fatal T . cruzi–T . dionisii mixed infection and occur in Triatoma vitticeps at the study site [32] . These DTUs were not detected in parasitological or serological tests on local rodents and marsupials [32] . Triatoma vitticeps is thought to have poor stercocarian vector competence [51] and oral transmission via insectivory may be one of the few ways in which this species propagates disease . The apparent transfer of trypanosome diversity en masse from bat to human host via ingestion of the vector [32] supports transmission efficiency reported elsewhere in oral outbreaks [52] . Furthermore , given the low terrestrial mammal abundance in the heavily fragmented region where the samples were collected [32] , bats may function here as principal reservoirs of parasites . There is growing evidence of bats’ potential in the maintenance of zoonotic T . cruzi transmission elsewhere in South America . For example , recent molecular surveys rank bats as top feeding sources of synanthropic T . cruzi-infected triatomines throughout Colombia , emphasize bats’ bridging of domestic and sylvatic transmission cycles in rural areas of Ecuador [45] ( where non-volant hosts have shown limited infection [53–54] ) and implicate bats as long-term refuges for parasites in areas subject to transmission interventions in Argentina [46] . Evidence of a new T . cruzi genotype associated with anthropogenic bats ( TcBat ) is also accumulating from around the continent [45 , 55–58] . TcBat was not , however , observed in this study . Here , we also provide first report of T . rangeli lineage D in bats , a strain initially isolated from Phyllomys dasythrix in southern Brazil [59] . As the ecogeographical structure of the Rhodnius spp . complex is thought to drive lineage divergence in T . rangeli [60–61] , an efficiently transmitted salivarian parasite , our detection of lineage D further north and beyond the Rodentia serves well to confirm theory . Its putative vector R . domesticus [62] occurs throughout the Atlantic Forest , often in bromeliads [63] that rely on nectarivorous bats ( e . g . , the specialist flower-feeder A . caudifer ) for pollination [64] . Whilst the expansion of the range of T . rangeli comes as little surprise , the presence of trypanosomes ( OTU 4 ) with reptilian affinities in our study population is perhaps more intriguing . Nonetheless , bats and reptiles do commonly co-occur in an arboreal niche . Ecological host-fitting , involving opportunistic host switching mediated by vectors' feeding patterns within an ecological niche , is thought to be a prevailing mode of trypanosome evolution [65] . Reptilian trypanosomes are transmitted by sand fly vectors [65–66] , with reports from Amazonia ( Viannamyia tuberculate [67] ) as well as central Brazil ( Evandromyia evandroi [68] ) . Shared microhabitat use among bats , reptiles and sand flies potentiates spill-over of the parasite . Most trypanosomatid diversity observed in this study was associated with complex mixed infections , a likely consequence of bats’ gregarious way of life . Tolerance of intracellular pathogens in the Chiroptera [21] suggests that multiple subclinical/asymptomatic infections may well accumulate in these hosts before triggering pathology linked to adaptive immune reactions in other non-volant mammals [69–72] . Frequent mixed infections , often coupled with low parasitaemia , have impeded bat trypanosome surveys in the past , both in genotyping from primary samples ( e . g . , low sensitivity in classic barcoding ) [73] and on cultured cells ( e . g . , growth bias ) [55 , 61 , 74] . The data presented here suggest that deep sequencing can resolve both infection identity and complexity . Although our study demonstrates the power of the metabarcoding approach , several caveats are relevant . Sensitivity to contamination and errors from amplification and sequencing are of foremost concern [27 , 75] . We employed a variety of cautionary measures during sample processing ( e . g . , flame-sterilized blood withdrawal , multiple negative DNA extraction/amplification controls ) and in the bioinformatic phase: prior to taxonomic inference , we sent sequenced amplicons through a severe quality filter ( 99 . 9% base call accuracy ) , absorbed potential artefactual variance into broad 98% similarity clusters and rejected unassigned OTUs present at low to moderate depth ( < 300/600 reads ) . Nevertheless , our study would have benefited from the inclusion of traditional methods ( e . g . , microscopy , ex and in vivo culture ) for validation and follow-up . Based on subunit rRNA , OTU 9 , isolated from a single bat ( D . rotundus ) , was assigned to B . saltans , considered the closest free-living relative of the parasitic trypanosomatids . This observation joins others in unsettling assumptions about putatively free-living , yet seldom studied protist taxa . For example , 18S rRNA analysis ( complemented by microscopy and serological testing ) found an apparent case of babesiosis in China to involve erythrocytic colpodellids , the closest “free-living” relatives of the parasitic Apicomplexa [76] . Regrettably , our field-based study passed over visual and biochemical tests that could have established the occurrence and viability of OTU 9 in mammalian tissue and we hesitate to entirely rule out environmental contamination as its source . Bodo saltans belongs to the most widely adapted , physiologically tolerant zooflagellates on Earth [77] . It abounds in soil and water and can also spread in aerosolized forms . As such , this eubodonid may in rare cases happen upon sampling equipment as well as resist certain antiseptic measures taken in the field . On account of its exceptional halotolerance [78] , for example , B . saltans may withstand some iodine-based disinfection ( as do other protozoans—e . g . , Cryptosporidium and Giardia [79] ) , though very unlikely as performed in this study ( i . e . , with ethanol ) . More importantly , however , OTU detection does not require a living organism , only its DNA . Severe contamination from the “dead” DNA of protist flagellates has indeed preoccupied past rRNA sequence analysis ( e . g . , see methods in [80] ) . In any case , we suggest additional ( environmental ) control samples ( e . g . , vials opened in the field , topical swabs around the site of cardiac puncture ) and laboratory efforts that distinguish DNA from viable cells ( e . g . , separation of lysed and non-lysed cells , RNA/DNA comparisons ) to help test for such possibilities in future research . In this section of Atlantic Forest , where a rural Chagas disease fatality in all likelihood involved a bat-feeding triatomine [32] , our deep sequencing study highlights the role of the Chiroptera as a reservoir for trypanosomiases . Furthermore , the unprecedented transfer of T . dionisii to a human from a bat , as well as the presence of reptile-infecting and putatively non-parasitic kinetoplastids in the same bat population , highlights the role of bats as keystone species in parasite spill-over events . Many questions remain on how the role of sylvatic hosts in pathogen dispersal varies in space and time , upon change to environment and at the evolutionary scale . Research into these intricacies of complex zoonosis will require much further innovation with high-sensitivity , high-throughput tools . We point to the power of next-generation metabarcoding strategies in studies of trypanosomatid ecology and evolution and strongly commend their future complementation with non-molecular methods .
|
Bats make up a mega-diverse , intensely parasitized order of volant mammals whose unique behavioural and physiological adaptations promote infection by a vast array of microorganisms . Trypanosomes stand out as ancient protozoan parasites of bats . As cryptic morphology , low parasitaemia and selective growth in culture have recurrently biased survey , we used 18S ribosomal RNA metabarcoding to resolve bat trypanosomatid diversity in Atlantic Forest fragments of southeast Brazil . Next to several unknown species , our deep sequence-based detection and assignment protocol recognized multiple known human-pathogenic trypanosomes , another linked to reptile hosts as well as a non-parasitic kinetoplastid in the blood of various phyllostomid bats . The striking permissivity exposed here , in a region where bat trypanosomes recently featured in a fatal case of Chagas disease , compels further research on bats’ role in the dispersal and spill-over of various microorganisms among humans and wildlife .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"geographical",
"locations",
"vertebrates",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"animals",
"mammals",
"protozoans",
"molecular",
"biology",
"techniques",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"south",
"america",
"artificial",
"gene",
"amplification",
"and",
"extension",
"pathogenesis",
"molecular",
"biology",
"brazil",
"ribosomes",
"people",
"and",
"places",
"trypanosoma",
"cruzi",
"biochemistry",
"rna",
"trypanosoma",
"ribosomal",
"rna",
"cell",
"biology",
"nucleic",
"acids",
"host-pathogen",
"interactions",
"polymerase",
"chain",
"reaction",
"bats",
"biology",
"and",
"life",
"sciences",
"non-coding",
"rna",
"amniotes",
"organisms"
] |
2017
|
Small subunit ribosomal metabarcoding reveals extraordinary trypanosomatid diversity in Brazilian bats
|
Contemporary theory of spiking neuronal networks is based on the linear response of the integrate-and-fire neuron model derived in the diffusion limit . We find that for non-zero synaptic weights , the response to transient inputs differs qualitatively from this approximation . The response is instantaneous rather than exhibiting low-pass characteristics , non-linearly dependent on the input amplitude , asymmetric for excitation and inhibition , and is promoted by a characteristic level of synaptic background noise . We show that at threshold the probability density of the potential drops to zero within the range of one synaptic weight and explain how this shapes the response . The novel mechanism is exhibited on the network level and is a generic property of pulse-coupled networks of threshold units .
Understanding the dynamics of single neurons , recurrent networks of neurons , and spike-timing dependent synaptic plasticity requires the quantification of how a single neuron transfers synaptic input into outgoing spiking activity . If the incoming activity has a slowly varying or constant rate , the membrane potential distribution of the neuron is quasi stationary and its steady state properties characterize how the input is mapped to the output rate . For fast transients in the input , time-dependent neural dynamics gains importance . The integrate-and-fire neuron model [1] can efficiently be simulated [2] , [3] and well approximates the properties of mammalian neurons [4]–[6] and more detailed models [7] . It captures the gross features of neural dynamics: The membrane potential is driven by synaptic impulses , each of which causes a small deflection that in the absence of further input relaxes back to a resting level . If the potential reaches a threshold , the neuron emits an action potential and the membrane potential is reset , mimicking the after-hyperpolarization . The analytical treatment of the threshold process is hampered by the pulsed nature of the input . A frequently applied approximation treats synaptic inputs in the diffusion limit , in which postsynaptic potentials are vanishingly small while their rate of arrival is high . In this limit , the summed input can be replaced by a Gaussian white noise current , which enables the application of Fokker-Planck theory [8] , [9] . For this approximation the stationary membrane potential distribution and the firing rate are known exactly [8] , [10] , [11] . The important effect of synaptic filtering has been studied in this limit as well; modelling synaptic currents as low-pass filtered Gaussian white noise with non-vanishing temporal correlations [12]–[15] . Again , these results are strictly valid only if the synaptic amplitudes tend to zero and their rate of arrival goes to infinity . For finite incoming synaptic events which are excitatory only , the steady state solution can still be obtained analytically [16] , [17] and also the transient solution can efficiently be obtained by numerical solution of a population equation [18] . A different approach takes into account non-zero synaptic amplitudes to first calculate the free membrane potential distribution and then obtain the firing rate by solving the first passage time problem numerically [19] . This approach may be extendable to conductance based synapses [20] . Exact results for the steady state have so far only been presented for the case of exponentially distributed synaptic amplitudes [21] . The spike threshold renders the model an extremely non-linear unit . However , if the synaptic input signal under consideration is small compared to the total synaptic barrage , a linear approximation captures the main characteristics of the evoked response . In this scenario all remaining inputs to the neuron are treated as background noise ( see Figure 1A ) . Calculations of the linear response kernel in the diffusion limit suggested that the integrate-and-fire model acts as a low-pass filter [22] . Here spectrum and amplitude of the synaptic background input are decisive for the transient properties of the integrate-and-fire model: in contrast to white noise , low-pass filtered synaptic noise leads to a fast response in the conserved linear term [12] . Linear response theory predicts an optimal level of noise that promotes the response [23] . In the framework of spike-response models , an immediate response depending on the temporal derivative of the postsynaptic potential has been demonstrated in the regime of low background noise [24] . The maximization of the input-output correlation at a finite amplitude of additional noise is called stochastic resonance and has been found experimentally in mechanoreceptors of crayfish [25] , in the cercal sensory system of crickets [26] , and in human muscle spindles [27] . The relevance and diversity of stochastic resonance in neurobiology was recently highlighted in a review article [28] . Linear response theory enables the characterization of the recurrent dynamics in random networks by a phase diagram [22] , [29] . It also yields approximations for the transmission of correlated activity by pairs of neurons in feed-forward networks [30] , [31] . Furthermore , spike-timing dependent synaptic plasticity is sensitive to correlations between the incoming synaptic spike train and the firing of the neuron ( see Figure 1 ) , captured up to first order by the linear response kernel [32]–[38] . For neuron models with non-linear membrane potential dynamics , the linear response properties [39] , [40] and the time-dependent dynamics can be obtained numerically [41] . Afferent synchronized activity , as it occurs e . g . in primary sensory cortex [42] , easily drives a neuron beyond the range of validity of the linear response . In order to understand transmission of correlated activity , the response of a neuron to fast transients with a multiple of a single synaptic amplitude [43] hence needs to be quantified . In simulations of neuron models with realistic amplitudes for the postsynaptic potentials , we observed a systematic deviation of the output spike rate and the membrane potential distribution from the predictions by the Fokker-Planck theory modeling synaptic currents by Gaussian white noise . We excluded any artifacts of the numerics by employing a dedicated high accuracy integration algorithm [44] , [45] . The novel theory developed here explains these observations and lead us to the discovery of a new early component in the response of the neuron model which linear response theory fails to predict . In order to quantify our observations , we extend the existing Fokker-Planck theory [46] and hereby obtain the mean time at which the membrane potential first reaches the threshold; the mean first-passage time . The advantage of the Fokker-Planck approach over alternative techniques has been demonstrated [47] . For non-Gaussian noise , however , the treatment of appropriate boundary conditions for the membrane potential distribution is of utmost importance [48] . In the results section we develop the Fokker-Planck formalism to treat an absorbing boundary ( the spiking threshold ) in the presence of non-zero jumps ( postsynaptic potentials ) . For the special case of simulated systems propagated in time steps , an analog theory has recently been published by the same authors [49] , which allows to assess artifacts introduced by time-discretization . Our theory applied to the integrate-and-fire model with small but finite synaptic amplitudes [1] , introduced in section “The leaky integrate-and-fire model” , quantitatively explains the deviations of the classical theory for Gaussian white noise input . After reviewing the diffusion approximation of a general first order stochastic differential equation we derive a novel boundary condition in section “Diffusion with finite increments and absorbing boundary” . We then demonstrate in section “Application to the leaky integrate-and-fire neuron” how the steady state properties of the model are influenced: the density just below threshold is increased and the firing rate is reduced , correcting the preexisting mean first-passage time solution [10] for the case of finite jumps . Turning to the dynamic properties , in section “Response to fast transients” we investigate the consequences for transient responses of the firing rate to a synaptic impulse . We find an instantaneous , non-linear response that is not captured by linear perturbation theory in the diffusion limit and that displays marked stochastic resonance . On the network level , we demonstrate in section “Dominance of the non-linear component on the network level” that the non-linear fast response becomes the most important component in case of feed-forward inhibition . In the discussion we consider the limitations of our approach , mention possible extensions and speculate about implications for neural processing and learning .
Consider a leaky integrate-and-fire model [1] with membrane time constant and resistance receiving excitatory and inhibitory synaptic inputs , as they occur in balanced neural networks [50] . We aim to obtain the mean firing rate and the steady state membrane potential distribution . The input current is modeled by point events , drawn from homogeneous Poisson processes with rates and , respectively . The membrane potential is governed by the differential equation . An excitatory spike causes a jump of the membrane potential by , an inhibitory spike by , so , where is a constant background current . Whenever reaches the threshold , the neuron emits a spike and the membrane potential is reset to , where it remains clamped for the absolute refractory time . The approach we take is to modify the existing Fokker-Planck theory in order to capture the major effects of the finite jumps . To this end , we derive a novel boundary condition at the firing threshold for the steady state membrane potential distribution of the neuron . We then solve the Fokker-Planck equation obtained from the standard diffusion approximation [8] , [10] , [11] , [22] , [23] given this new condition .
The membrane potential of the model neuron follows a first order stochastic differential equation . Therefore , in this section we consider a general first order stochastic differential equation driven by point events . In order to distinguish the dimensionless quantities in this section from their counterparts in the leaky integrate-and-fire model , we denote the rates of the two incoming Poisson processes by ( excitation ) and ( inhibition ) . Each incoming event causes a finite jump ( the excitatory synaptic weight ) for an increasing event and ( the inhibitory synaptic weight ) for a decreasing event . The stochastic differential equation takes the form ( 1 ) where captures the deterministic time evolution of the system ( with for the leaky integrate-and-fire neuron ) . We follow the notation in [46] and employ the Kramers-Moyal expansion with the infinitesimal moments . The first and second infinitesimal moment evaluate to and , where we introduced the shorthand and . The time evolution of the probability density is then governed by the Kramers-Moyal expansion , which we truncate after the second term to obtain the Fokker-Planck equation ( 2 ) where denotes the probability flux operator . In the presence of an absorbing boundary at , we need to determine the resulting boundary condition for the stationary solution of ( 2 ) . Without loss of generality , we assume the absorbing boundary at to be the right end of the domain . A stationary solution exists , if the probability flux exiting at the absorbing boundary is reinserted into the system . For the example of an integrate-and-fire neuron , reinsertion takes place due to resetting the neuron to the same potential after each threshold crossing . This implies a constant flux through the system between the point of insertion and threshold . Rescaling the density by this flux as results in the stationary Focker-Planck equation , which is a linear inhomogeneous differential equation of first order ( 3 ) with . First we consider the diffusion limit , in which the rate of incoming events diverges , while the amplitude of jumps goes to zero , such that mean and fluctuations remain constant . In this limit , the Kramers-Moyal expansion truncated after the second term becomes exact [51] . This route has been taken before by several authors [8] , [22] , [23] , here we review these results to consistently present our extension of the theory . In the above limit equation ( 3 ) needs to be solved with the boundary conditionsMoreover , a finite probability flux demands the density to be a continuous function , because of the derivative in the flux operator . In particular , the solution must be continuous at the point of flux insertion ( however , the first derivative is non-continuous at due to the step function in the right hand side of ( 3 ) ) . Continuity especially implies a vanishing density at threshold . Once the solution of ( 3 ) is found , the normalization condition determines the stationary flux . Now we return to the problem of finite jumps . We proceed along the same lines as in the diffusion limit , seeking the stationary solution of the Fokker-Planck equation ( 2 ) . We keep the boundary conditions at and at as well as the normalization condition as before , but we need to find a new self-consistent condition at threshold , because the density does not necessarily have to vanish if the rate of incoming jumps is finite . The main assumption of our work is that the steady state solution satisfies the stationary Fokker-Planck equation ( 3 ) based on the diffusion approximation within the interval , but not necessarily at the absorbing boundary , where the solution might be non-continuous . To obtain the boundary condition , we note that the flux over the threshold has two contributions , the deterministic drift and the positive stochastic jumps crossing the boundary ( 4 ) ( 5 ) with . To evaluate the integral in ( 5 ) , for small we expand into a Taylor series around . This is where our main assumption enters: we assume that the stationary Fokker-Planck equation ( 3 ) for is a sufficiently accurate characterization of the jump diffusion process . We solve this equation for It is easy to see by induction , that the function and all its higher derivatives , can be written in the form , whose coefficients for obey the recurrence relation ( 6 ) with the additional values and , as denotes the function itself . Inserting the Taylor series into ( 5 ) and performing the integration results in ( 7 ) which is the probability mass moved across threshold by a perturbation of size and hence also quantifies the instantaneous response of the system . After dividing ( 4 ) by we solve for to obtain the Dirichlet boundary condition ( 8 ) If is small compared to the length scale on which the probability density function varies , the probability density near the threshold is well approximated by a Taylor polynomial of low degree; throughout this work , we truncate ( 7 ) and ( 12 ) at . The boundary condition ( 8 ) is consistent with in the diffusion limit , in which the rate of incoming jumps diverges , while their amplitude goes to zero , such that the first ( ) and second moment ( ) stay finite . This can be seen by scaling , , with such that the mean is kept constant [51] . Inserting this limit in ( 8 ) , we find ( 9 ) since , and vanishes for , is bounded and . The general solution of the stationary Fokker-Planck equation ( 3 ) is a sum of a homogeneous solution that satisfies and a particular solution with . The homogeneous solution is , where we fixed the integration constant by chosing . The particular solution can be obtained by variation of constants and we chose it to vanish at the threshold as . The complete solution is a linear combination , where the prefactor is determined by the boundary condition ( 8 ) in the case of finite jumps , or by for Gaussian white noise The normalization condition determines the as yet unknown constant probability flux through the system . We now apply the theory developed in the previous section to the leaky integrate-and-fire neuron with finite postsynaptic potentials . Due to synaptic impulses , the membrane potential drifts towards and fluctuates with the diffusion constant . This suggests to choose the natural units for the time and for the voltage to obtain the simple expressions for the drift- and for the diffusion-term in the Fokker-Planck operator ( 2 ) . The probability flux operator ( 2 ) is then given as . In the same units the stationary probability density scaled by the flux reads where is the flux corresponding to the firing rate in units of . As is already scaled by the flux , application of the flux operator yields unity between reset and threshold and zero outside ( 10 ) The steady state solution of this stationary Fokker-Planck equation ( 11 ) is a linear superposition of the homogeneous solution and the particular solution . The latter is chosen to be continuous at and to vanish at . Using the recurrence ( 6 ) for the coeffcients of the Taylor expansion of the membrane potential density , we obtain and , where starts from . The first important result of this section is the boundary value of the density at the threshold following from ( 8 ) as ( 12 ) The constant in ( 11 ) follows from . The second result is the steady state firing rate of the neuron . With being the fraction of neurons which are currently refractory , we obtain the rate from the normalization condition of the density as ( 13 ) The normalized steady state solution Figure 2A therefore has the complete form ( 14 ) Figure 2B , D shows the steady state solution near the threshold obtained by direct simulation to agree much better with our analytical approximation than with the theory for Gaussian white noise input . Even for synaptic amplitudes ( here ) which are considerably smaller than the noise fluctuations ( here ) , the effect is still well visible . The oscillatory deviations with periodicity close to reset observable in Figure 2A are due to the higher occupation probability of voltages that are integer multiples of a synaptic jump away from reset . The modulation washes out due to coupling of adjacent voltages by the deterministic drift as one moves away from reset . The oscillations at lower frequencies apparent in Figure 2A are due to aliasing caused by the finite bin width of the histogram ( ) . The synaptic weight is typically small compared to the length scale on which the probability density function varies . So the probability density near the threshold is well approximated by a Taylor polynomial of low degree; throughout this work , we truncate the series in ( 12 ) at . A comparison of this approximation to the full solution is shown in Figure 2E . For small synaptic amplitudes ( shown ) , below threshold and outside the reset region ( Figure 2A , C ) the approximation agrees with the simulation within its fluctuation . At the threshold ( Figure 2B , D ) our analytical solution assumes a finite value whereas the direct simulation only drops to zero on a very short voltage scale on the order of the synaptic amplitude . For larger synaptic weights ( , see Figure 2F ) , the density obtained from direct simulation exhibits a modulation on the corresponding scale . The reason is the rectifying nature of the absorbing boundary: A positive fluctuation easily leads to a threshold crossing and absorption of the state in contrast to negative fluctuations . Effectively , this results in a net drift to lower voltages within the width of the jump distribution caused by synaptic input , visible as the depletion of density directly below the threshold and an accumulation further away , as observed in Figure 2F . The second term ( proportional to ) appearing in ( 13 ) is a correction to the well known firing rate equation of the integrate-and-fire model driven by Gaussian white noise [10] . Figure 3 compares the firing rate predicted by the new theory to direct simulation and to the classical theory . The classical theory consistently overestimates the firing rate , while our theory yields better accuracy . Our correction resulting from the new boundary condition becomes visible at moderate firing rates when the density slightly below threshold is sufficiently high . At low mean firing rates , the truncation of the Kramers-Moyal expansion employed in the Fokker-Planck description may contribute comparably to the error . Our approximation captures the dependence on the synaptic amplitude correctly for synaptic amplitudes of up to ( Figure 3B ) . The insets in Figure 3C , D show the relative error of the firing rate as a function of the noise amplitude . As expected , the error increases with the ratio of the synaptic effect compared to the amplitude of the noise fluctuations . For low noise , our theory reduces the relative error by a factor of compared to the classical diffusion approximation . We now proceed to obtain the response of the firing rate to an additional -shaped input current . Such a current can be due to a single synaptic event or due to the synchronized arrival of several synaptic pulses . In the latter case , the effective amplitude of the summed inputs can easily exceed that of a single synapse . The fast current transient causes a jump of the membrane potential at and ( 2 ) suggests to treat the incident as a time dependent perturbation of the mean input . First , we are interested in the integral response of the excess firing rate . Since the perturbation has a flat spectrum , up to linear order in the spectrum of the excess rate is , where is the linear transfer function with respect to perturbing at Laplace frequency . In particular , . As is the DC susceptibility of the system , we can express it up to linear order as . Hence , ( 15 ) We also take into account the dependence of on to calculate from ( 13 ) and obtain ( 16 ) Figure 4D shows the integral response to be in good agreement with the linear approximation . This expression is consistent with the result in the diffusion limit : Here the last term becomes , where we used , following from ( 10 ) with . This results in , which can equivalently be obtained directly as the derivative of ( 13 ) with respect to setting . Taking the limit , however , does not change significantly the integral response compared to the case of finite synaptic amplitudes ( Figure 4D , Figure 5A ) . The instantaneous response of the firing rate to an impulse-like perturbation can be quantified without further approximation . The perturbation shifts the probability density by so that neurons with ] immediately fire . This results in the finite firing probability of the single neuron within infinitesimal time ( 5 ) , which is zero for . This instantaneous response has several interesting properties: For small it can be approximated in terms of the value and the slope of the membrane potential distribution below the threshold ( using ( 7 ) for ) , so it has a linear and a quadratic contribution in . Figure 4A shows a typical response of the firing rate to a perturbation . The peak value for a positive perturbation agrees well with the analytical approximation ( 7 ) ( Figure 4C ) . Even in the diffusion limit , replacing the background input by Gaussian white noise , the instantaneous response persists . Using the boundary condition our theory is applicable to this case as well . Since the density just below threshold is reduced , ( 5 ) yields a smaller instantaneous response ( Figure 4C , Figure 5B ) which for positive still exhibits a quadratic , but no linear , dependence . The increasing and convex dependence of the response probability on the amplitude of the perturbation is a generic feature of neurons with subthreshold mean input that also persists in the case of finite synaptic rise time . In this regime , the membrane potential distribution has a mono-modal shape centered around the mean input , which is inherited from the underlying superposition of a large number of small synaptic impulses . The decay of the density towards the threshold is further enhanced by the probability flux over the threshold: a positive synaptic fluctuation easily leads to the emission of a spike and therefore to the absorption of the state at the threshold , depleting the density there . Consequently , the response probability of the neuron is increasing and convex as long as the peak amplitude of the postsynaptic potential is smaller than the distance of the peak of the density to the threshold . It is increasing and concave beyond this point . At present the integrate-and-fire model is the simplest analytically tractable model with this feature . The integral response ( 15 ) as well as the instantaneous response ( 5 ) both exhibit stochastic resonance; an optimal level of synaptic background noise enhances the transient . Figure 5A shows this noise level to be at about for the integral response . The responses to positive and negative perturbations are symmetric and the maximum is relatively broad . The instantaneous response in Figure 5B displays a pronounced peak at a similar value of . This non-linear response only exists for positive perturbations; the response is zero for negative ones . Though the amplitude is reduced in the case of Gaussian white noise background , the behavior is qualitatively the same as for noise with finite jumps . Stochastic resonance has been reported for the linear response to sinusoidal periodic stimulation [23] . Also for non-periodic signals that are slow compared to the neuron's dynamics an adiabatic approximation reveals stochastic resonance [52] . In contrast to the latter study , the rate transient observed in our work is the instantaneous response to a fast ( Dirac ) synaptic current . Due to the convex nature of the instantaneous response ( Figure 4C ) its relative contribution to the integral response increases with . For realistic synaptic weights the contribution reaches percent . An example network in which the linear non-instantaneous response cancels completely and the instantaneous response becomes dominant is shown in Figure 6A . At two populations of neurons simultaneously receive a perturbation of size and respectively . This activity may , for example , originate from a third pool of synchronous excitatory and inhibitory neurons . It may thus be interpreted as feed-forward inhibition . The linear contributions to the pooled firing rate response of the former two populations hence is zero . The instantaneous response , however , causes a very brief overshoot at ( Figure 6B ) . Figure 6C reveals that the response returns to baseline within . Figure 6D shows that the dependence of peak height on still exhibits the supra-linearity . The quite exact cancellation of the response for originates from the symmetry of the response functions for positive and negative perturbations in this interval ( shown in Figure 4A , B ) . The pooled firing rate of the network is the sum of the full responses: the instantaneous response at does not share the symmetry and hence does not cancel . This demonstrates that the result of linear perturbation theory is a good approximation for and that the instantaneous response at the single time point completes the characterization of the neuronal response .
In this work we investigate the effect of small , but non-zero synaptic impulses on the steady state and response properties of the integrate-and-fire neuron model . We obtain a more accurate description of the firing rate and the membrane potential distribution in the steady state than provided by the classical approximation of Gaussian white noise input currents [10] . Technically this is achieved by a novel hybrid approach combining a diffusive description of the membrane potential dynamics far away from the spiking threshold with an explicit treatment of threshold crossings by synaptic transients . This allows us to obtain a boundary condition for the membrane potential density at threshold that captures the observed elevation of density . Our work demonstrates that in addition to synaptic filtering , the granularity of the noise due to finite non-zero amplitudes does affect the steady state and the transient response properties of the neuron . Here , we study the effect of granularity using the example of a simple neuron model with only one dynamic variable . The quantitatively similar increase of the density close to threshold observed if low-pass filtered Gaussian white noise is used as a model for the synaptic current has a different origin . It is due to the absence of a diffusion term in the dynamics of the membrane potential [12] , [13] , [15] . The analytical treatment of finite synaptic amplitudes further allows us to characterize the probability of spike emission in response to synaptic inputs for neuron models with a single dynamical variable and renewal . Alternatively , this response can be obtained numerically from population descriptions [18] , [39]–[41] or , for models with one or more dynamic variables and gradually changing inputs , in the framework of the refractory density approximation [15] . Here , we find that the response can be decomposed into a fast , non-linear and a slow linear contribution , as observed experimentally about a quarter of a century ago [53] in motor neurons of cat cortex in the presence of background noise . The existence of a fast contribution proportional to the temporal change of the membrane potential was predicted theoretically [54] . In the framework of the refractory density approach [15] , the effective hazard function of an integrate-and-fire neuron also exhibits contributions to spike emission due to two distinct causes: the diffusive flow through the threshold and the movement of density towards the threshold . The latter contribution is proportional to the temporal change of the membrane potential and is corresponding to the instantaneous response reported here , but for the case of a gradually increasing membrane potential . Contemporary theory of recurrent networks so far has neglected the transient non-linear component of the neural response , an experimentally observed feature [53] that is generic to threshold units in the presence of noise . The infinitely fast rise of the postsynaptic potential in the integrate-and-fire model leads to the immediate emission of a spike with finite probability . For excitatory inputs , this probability depends supra-linearly on the amplitude of the synaptic impulse and it is zero for inhibitory impulses . The supra-linear increase for small positive impulse amplitudes relates to the fact that the membrane potential density decreases towards threshold: the probability to instantaneously emit a spike equals the integral of the density shifted over the threshold . The detailed shape of the density below threshold therefore determines the response properties . For Gaussian white noise synaptic background , the model still displays an instantaneous response . However , since in this case the density vanishes at threshold , the response probability to lowest order grows quadratically in the amplitude of a synaptic impulse . This is the reason why previous work based on linear response theory did not report on the existence of an instantaneous component when modulating the mean input and on the contrary characterized the nerve cell as a low-pass in this case [22] , [23] . Modulation of the noise amplitude , however , has been shown to cause an instantaneous response in linear approximation in the diffusion limit [23] , confirmed experimentally in real neurons [55] . While linear response theory has proven extremely useful to understand recurrent neural networks [29] , the categorization of the integrate-and-fire neuron's response kernel as a low-pass is misleading , because it suggests the absence of an immediate response . Furthermore we find that in addition to the nature of the background noise , response properties also depend on its amplitude: a certain level of noise optimally promotes the spiking response . Hence noise facilitates the transmission of the input to the output of the neuron . This is stochastic resonance in the general sense of the term as recently suggested [28] . As noted in the introduction , stochastic resonance of the linear response kernel has previously been demonstrated for sinusoidal input currents and Gaussian white background noise [23] . Furthermore , also slow aperiodic transients are facilitated by stochastic resonance in the integrate-and-fire neuron [52] . We extend the known results in two respects . Firstly , we show that the linear response shows aperiodic stochastic resonance also for fast transients . Secondly , we demonstrate that the instantaneous non-linear response exhibits a qualitatively similar , but even more pronounced dependence on noise intensity . For realistically small synaptic amplitudes , the instantaneous non-linear response is typically small compared to the linear contribution . However , this changes at the network level in the presence of feed-forward inhibition: a synchronized pair of an excitatory and an inhibitory pulse evokes spiking responses in two distinct neural populations , whose linear contributions mutually cancel and only the non-linear immediate contribution remains . Hence the immediate response dominates even for small synaptic amplitudes . The presented approximate analytical results are illustrated and confirmed by direct simulation . The instantaneous non-linear response is potentially a relevant mechanism for processing of transient signals by neurons . In auditory cortex , the irregular firing of neurons has been shown to be driven by simultaneous coactivation of several of their synaptic afferents [56] . The effective postsynaptic potential hence has the amplitude of multiple single synapses , which easily drives the spiking response into the supra-linear regime . The convex increase of firing probability is of advantage to obtain output spikes closely locked to the input . Furthermore , the non-linearity enables the neuron to perform non-trivial computations on the inputs [57] . In particular the memory capacity of networks in a categorization task can be increased by non-linear elements [58] . The circuit presented in section “Dominance of the non-linear component at the network level” establishes a quadratic input-output relationship for fast transient signals that may be useful for non-linear processing , analogous to the non-linear f-I curve ( spike frequency as a function of input current ) in the case of quasi-stationary rate-coded signals . Our finding of an immediate non-linear response has an implication on the intensely debated question how common input affects the correlation of the spiking activity of pairs of neurons [30] , [31] , [59] . The immediate response adds to the correlation at zero time lag , because it increases the probability of both neurons to simultaneously emit a spike . Due to the non-linearity of the mechanism , the immediate firing probability easily becomes the dominant contribution . Our theory yields a means to quantitatively assess this contribution to firing synchrony . Synapses with spike timing dependent plasticity ( STDP ) [60] are sensitive to the input-spike triggered firing rate of the neuron . The fast response is relevant , because closely time-locked pre- and postsynaptic activity most effectively changes the synaptic weight . This is illustrated in Figure 1B . The direction of weight change depends on whether the fast response falls on the potentiating or the depotentiating part of the STDP curve , determined by the difference between dendritic and axonal synaptic delay [33] . Assuming that the causal fast response strengthens the synapse ( Hebbian learning [61] ) , the supra-linearity combined with multiplicative spike-timing-dependent learning rules may add new fixed points for the synaptic weight and thus influence pattern formation in recurrent networks [62] . Previous work restricted the analysis of the interplay of neural dynamics and synaptic plasticity in feed-forward [32]–[34] as well as in recurrent networks [35]–[38] to the linear response of the neuron . Our framework extends the scope of analytical investigations of synaptic dynamics to the inherently non-linear response properties of neurons . The pronounced stochastic resonance of the individual neuron implies an optimal level of synaptic background noise that supports cooperativity among afferent synapses and hence also the sensitivity to correlations among them [34] . Measuring synaptic plasticity in the presence of network activity might elucidate how stochastic resonance influences cooperative synaptic learning . Postsynaptic potentials exhibit a finite rise time , whereas the membrane potential of the integrate-and-fire neuron model jumps at each incoming synaptic event . Although this is a simplification , the model reproduces experimental spike trains surprisingly well [6] . For non-zero rise times , the instantaneous firing rate response reported here is spread out in time over the rising flank of the postsynaptic potential and is proportional to the derivative of the membrane potential [43] , [54] . The asymmetry for excitatory and inhibitory synaptic events and the supralinear increase of the response probability with excitatory postsynaptic amplitude , however , are generic features that carry over to finite rise time if the neuron operates in the fluctuation driven regime . Comparing the non-linear and the linear response probability experimentally can serve as an indicator to decide on the importance of each contribution in real neurons . The integral linear response can be obtained from similar arguments as in section “Response to fast transients” as with the slope of the f-I curve and the membrane resistance . Previous work has shown that the spike generation mechanism influences the transient properties of neurons [63] , [64] . Specifically , a soft threshold , as realized in the exponential integrate-and-fire neuron model [63] is more realistic than the hard threshold of the leaky integrate-and-fire model considered here . Future work needs to investigate how this affects the fast response . We expect qualitatively similar findings , because a positive synaptic impulse shifts membrane potential density into the basin of attraction for spike generation . This will then result in an increased spiking density in a finite time window following the synaptic event . The hybrid approach combining a diffusion approximation with an explicit treatment of finite jumps near the boundary allowed us to uncover hitherto unknown properties of the integrate-and-fire model by analytical means . The diffusion approximation , however , still limits our approach: for synaptic amplitudes moments of order higher than two , which are neglected by the Fokker-Planck equation , become relevant . A combination of our boundary condition with an assessment of higher moments [19] , [65] seems promising . Also , the oscillatory modulations of the probability density on a scale in the regions below threshold and around the reset potential are outside the scope of our theory . The response properties considered in this work are entirely based on the assumption , that the dynamics has reached the steady state prior to arrival of the perturbing input . A valuable future extension of our work is to consider finite amplitude synaptic background noise and additional sinusoidal current injection . This would allow to quantify in a frequency resolved manner how the transfer properties of the model are influenced by finite-grained noise . Technically , the linear perturbation theory for the diffusion limit [22] would have to be combined with our boundary condition . Complications might arise from the fact that the boundary condition is now time-dependent if the mean drive reaches suprathreshold values in certain epochs . Our treatment of stochastic differential equations with finite jumps and absorbing boundaries is general , as long as the jumps are sufficiently small . We expect it to be applicable to other fluctuation driven dynamical systems in quantitative biology and physics . Potential areas include the diffusion of particles in domains with absorbing walls , chemical reactions with activation thresholds , circuit theory and solid state physics .
|
Our work demonstrates a fast-firing response of nerve cells that remained unconsidered in network analysis , because it is inaccessible by the otherwise successful linear response theory . For the sake of analytic tractability , this theory assumes infinitesimally weak synaptic coupling . However , realistic synaptic impulses cause a measurable deflection of the membrane potential . Here we quantify the effect of this pulse-coupling on the firing rate and the membrane-potential distribution . We demonstrate how the postsynaptic potentials give rise to a fast , non-linear rate transient present for excitatory , but not for inhibitory , inputs . It is particularly pronounced in the presence of a characteristic level of synaptic background noise . We show that feed-forward inhibition enhances the fast response on the network level . This enables a mode of information processing based on short-lived activity transients . Moreover , the non-linear neural response appears on a time scale that critically interacts with spike-timing dependent synaptic plasticity rules . Our results are derived for biologically realistic synaptic amplitudes , but also extend earlier work based on Gaussian white noise . The novel theoretical framework is generically applicable to any threshold unit governed by a stochastic differential equation driven by finite jumps . Therefore , our results are relevant for a wide range of biological , physical , and technical systems .
|
[
"Abstract",
"Introduction",
"Model",
"Results",
"Discussion"
] |
[
"biophysics/theory",
"and",
"simulation",
"neuroscience/theoretical",
"neuroscience",
"computational",
"biology/computational",
"neuroscience"
] |
2010
|
Instantaneous Non-Linear Processing by Pulse-Coupled Threshold Units
|
Animals repeat rewarded behaviors , but the physiological basis of reward-based learning has only been partially elucidated . On one hand , experimental evidence shows that the neuromodulator dopamine carries information about rewards and affects synaptic plasticity . On the other hand , the theory of reinforcement learning provides a framework for reward-based learning . Recent models of reward-modulated spike-timing-dependent plasticity have made first steps towards bridging the gap between the two approaches , but faced two problems . First , reinforcement learning is typically formulated in a discrete framework , ill-adapted to the description of natural situations . Second , biologically plausible models of reward-modulated spike-timing-dependent plasticity require precise calculation of the reward prediction error , yet it remains to be shown how this can be computed by neurons . Here we propose a solution to these problems by extending the continuous temporal difference ( TD ) learning of Doya ( 2000 ) to the case of spiking neurons in an actor-critic network operating in continuous time , and with continuous state and action representations . In our model , the critic learns to predict expected future rewards in real time . Its activity , together with actual rewards , conditions the delivery of a neuromodulatory TD signal to itself and to the actor , which is responsible for action choice . In simulations , we show that such an architecture can solve a Morris water-maze-like navigation task , in a number of trials consistent with reported animal performance . We also use our model to solve the acrobot and the cartpole problems , two complex motor control tasks . Our model provides a plausible way of computing reward prediction error in the brain . Moreover , the analytically derived learning rule is consistent with experimental evidence for dopamine-modulated spike-timing-dependent plasticity .
Many instances of animal behavior learning such as path finding in foraging , or – a more artificial example – navigating the Morris water-maze , can be interpreted as exploration and trial-and-error learning . In both examples , the behavior eventually learned by the animal is the one that led to high reward . These can be appetite rewards ( i . e . , food ) or more indirect rewards , such as the relief of finding the platform in the water-maze . Important progress has been made in understanding how learning of such behaviors takes place in the mammalian brain . On one hand , the framework of reinforcement learning [1] provides a theory and algorithms for learning with sparse rewarding events . A particularly attractive formulation of reinforcement learning is temporal difference ( TD ) learning [2] . In the standard setting , this theory assumes that an agent moves between states in its environment by choosing appropriate actions in discrete time steps . Rewards are given in certain conjunctions of states and actions , and the agent's aim is to choose its actions so as to maximize the amount of reward it receives . Several algorithms have been developed to solve this standard formulation of the problem , and some of these have been used with spiking neural systems . These include REINFORCE [3] , [4] and partially observable Markov decision processes [5] , [6] , in case the agent has incomplete knowledge of its state . On the other hand , experiments show that dopamine , a neurotransmitter associated with pleasure , is released in the brain when reward , or a reward-predicting event , occurs [7] . Dopamine has been shown to modulate the induction of plasticity in timing non-specific protocols [8]–[11] . Dopamine has also recently been shown to modulate spike-timing-dependent plasticity ( STDP ) , although the exact spike-timing and dopamine requirements for induction of long-term potentiation ( LTP ) and long-term depression ( LTD ) are still unclear [12]–[14] . A crucial problem in linking biological neural networks and reinforcement learning is that typical formulations of reinforcement learning rely on discrete descriptions of states , actions and time , while spiking neurons evolve naturally in continuous time and biologically plausible “time-steps” are difficult to envision . Earlier studies suggested that an external reset [15] or theta oscillations [16] might be involved , but no evidence exists to support this and it is not clear why evolution would favor slower decision steps over a continuous decision mechanism . Indeed biological decision making is often modeled by an integrative process in continuous time [17] , where the actual decision is triggered when the integrated value reaches a threshold . In this study , we propose a way to narrow the conceptual gap between reinforcement learning models and the family of spike-timing-dependent synaptic learning rules by using continuous representations of state , actions and time , and by deriving biologically plausible synaptic learning rules . More precisely , we use a variation of the Actor-Critic architecture [1] , [18] for TD learning . Starting from the continuous TD formulation by Doya [19] , we derive reward-modulated STDP learning rules which enable a network of model spiking neurons to efficiently solve navigation and motor control tasks , with continuous state , action and time representations . This can be seen as an extension of earlier works [20] , [21] to continuous actions , continuous time and spiking neurons . We show that such a system has a performance on par with that of real animals and that it offers new insight into synaptic plasticity under the influence of neuromodulators such as dopamine .
The goal of a reinforcement learning agent is to maximize its future rewards . Following Doya [10] , we define the continuous-time value function as ( 3 ) where the brackets represent the expectation over all future trajectories and future action choices , dependent on the policy . The parameter represents the reward discount time constant , analogous to the discount factor of discrete reinforcement learning . Its effect is to make rewards in the near future more attractive than distant ones . Typical values of for a task such as the water-maze task would be on the order of a few seconds . Eq . 3 represents the total quantity of discounted reward that an agent in position at time and following policy can expect . The policy should be chosen such that is maximized for all locations . Taking the derivative of Eq . 3 with respect to time yields the self-consistency equation [19] ( 4 ) Calculating requires knowledge of the reward function and of the environment dynamics ( Eq 1 ) . These are , however , unknown to the agent . Typically , the best an agent can do is to maintain a parametric estimator of the “true” value function . This estimator being imperfect , it is not guaranteed to satisfy Eq . 4 . Instead , the temporal difference error is defined as the mismatch in the self-consistency , ( 5 ) This is analog to the discrete TD error [1] , [19] ( 6 ) where the reward discount factor plays a role similar to the reward discount time constant . More precisely , for short steps , [19] . An estimator can be said to be a good approximation to if the TD error is close to zero for all . This suggests a simple way to learn a value function estimator: by a gradient descent on the squared TD error in the following way ( 7 ) where is a learning rate parameter and is the set of parameters ( synaptic weights ) that control the estimator of the value function . This approach , dubbed residual gradient [19] , [25] , [26] , yields a learning rule that is formally correct , but in our case suffers from a noise bias , as shown in Models . Instead , we use a different learning rule , suggested for the discrete case by Sutton and Barto [1] . Translated in a continuous framework , the aim of their optimization approach is that the value function approximation should match the true value function . This is equivalent to minimizing an objective function ( 8 ) A gradient descent learning rule on yields ( 9 ) Of course , because is unknown , this is not a particularly useful learning rule . On the other hand , using Eq . 4 , this becomes ( 10 ) where we merged into the learning rate without loss of generality . In the last step , we replaced the real value function derivative with its estimate , i . e . , , and then used the definition of from Eq . 5 . The substitution of by in Eq . 10 is an approximation , and there is in general no guarantee that the two values are similar . However the form of the resulting learning rule suggests it goes in the direction of reducing the TD error . For example , if is positive at time , updating the parameters in the direction suggested by Eq . 10 , will increase the value of , and thus decrease . In [19] , a heuristic shortcut was used to go directly from the residual gradient ( Eq . 7 ) to Eq . 10 . As noted by Doya [19] , the form of the learning rule in Eq . 10 is a continuous version of the discrete [1] , [27] with function approximation ( here with ) . This has been shown to converge with probability 1 [28] , [29] , even in the case of infinite ( but countable ) state space . This must be the case also for arbitrarily small time steps ( such as the finite steps usually used in computer simulations of a continuous system [19] ) , and thus it seems reasonable to expect that the continuous version also converges under reasonable assumptions , even though to date no proof exists . An important problem in reinforcement learning is the concept of temporal credit assignment , i . e . , how to propagate information about rewards back in time . In the framework of TD learning , this means propagating the TD error at time so that the value function at earlier times is updated in consequence . The learning rule Eq . 10 does not by itself offer a solution to this problem , because the expression of explicitly refers only to and at time . Therefore does not convey information about other times and minimizing does not a priori affect values and . This is in contrast to the discrete version of the TD error ( Eq . 6 ) , where the expression of explicitly links to and thus the TD error is back-propagated during subsequent learning trials . If , however , one assumes that the value function is continuous and continuously differentiable , changing the values of and implies changing the values of these functions in a finite vicinity of . This is in particular the case if one uses a parametric form for , in the form of a weighted mixture of smooth kernels ( as we do here , see next section ) . Therefore , the conjunction of a function approximation of the value function in the form of a linear combination of smooth kernels ensures that the TD error is propagated in time in the continuous case , allowing the temporal credit assignment problem to be solved . We now take the above derivation a step further by assuming that the value function estimation is performed by a spiking neuron with firing rate . A natural way of doing this is ( 11 ) where is the value corresponding to no spiking activity and is a scaling factor with units of [reward units]×s . A choice of enables negative values , despite the fact that the rate is always positive . We call this neuron a critic neuron , because its role is to maintain an estimate of the value function . Several aspects should be discussed at this point . Firstly , since the value function in Eq . 11 must depend on the state of the agent , we must assume that the neuron receives some meaningful synaptic input about the state of the agent . In the following we make the assumption that this input is feed-forward from the place cells to the ( spiking ) critic neuron . Secondly , while the value function is in theory a function only of the state at time , a spiking neuron implementation ( such as the simplified model we use here , see Models ) will reflect the recent past , in a manner determined by the shape of the excitatory postsynaptic potentials ( EPSP ) it receives . This is a limitation shared by all neural circuits processing sensory input with finite synaptic delays . In the rest of this study , we assume that the evolution of the state of the agent is slow compared to the width of an EPSP . In that limit , the firing rate of a critic neuron at time actually reflects the position of the agent at that time . Thirdly , the firing rate of a single spike-firing neuron is itself a vague concept and multiple definitions are possible . Let's start from its spike train ( where is the set of the neuron's spike times and is the Dirac delta , not to be confused with the TD signal ) . The expectation is a statistical average of the neuron's firing over many repetitions . It is the theoretically favored definition of the firing rate , but in practice it is not available in single trials in a biologically plausible setting . Instead , a common workaround is to use a temporal average , for example by filtering the spike train with a kernel ( 12 ) Essentially , this amounts to a trade-off between temporal accuracy and smoothness of the rate function , of which extreme cases are respectively the spike train ( extreme temporal accuracy ) and a simple spike count over a long time window with smooth borders ( no temporal information , extreme smoothness ) . In choosing a kernel , it should hold that , so that each spike is counted once , and one often wishes the kernel to be causal ( ) , so that the current firing rate is fully determined by past spike times and independent of future spikes . Another common approximation for the firing rate of a neuron consists in replacing the statistical average by a population average , over many neurons encoding the same value . Provided they are statistically independent of each other ( for example if the neurons are not directly connected ) , averaging their responses over a single trial is equivalent to averaging the responses of a single neuron over the same number of trials . Here we combine temporal and population averaging , redefining the value function as an average firing rate of neurons ( 13 ) where the instantaneous firing rate of neuron is defined by Eq . 12 , using its spike train and a kernel defined by ( 14 ) This kernel rises with a time constant and decays to 0 with time constant . One advantage of the definition of Eq . 12 is that the derivative of the firing rate of neuron with respect to time is simply ( 15 ) so that computing the derivative of the firing rate is simply a matter of filtering the spike train with the derivative of the kernel . This way , the TD error of Eq . 5 can be expressed as ( 16 ) where , again , denotes the spike train of neuron in the pool of critic neurons . Suppose that feed-forward weights lead from a state-representation neuron to neuron in the population of critic neurons . Can the critic neurons learn to approximate the value function by changing the synaptic weights ? An answer to this question is obtained by combining Eq . 10 with Eqs 13 and 16 , leading to a weights update ( 17 ) where is the time course of an EPSP and is the spike train of the presynaptic neuron , restricted to the spikes posterior to the last spike time of postsynaptic neuron . For simplicity , we merged all constants into a new learning rate . A more formal derivation can be found in Models . Let us now have a closer look at the shape of the learning rule suggested by Eq . 17 . The effective learning rate is given by a parameter . The rest of the learning rule consists of a product of two terms . The first one is the TD error term , which is the same for all synapses , and can thus be considered as a global factor , possibly transmitted by one or more neuromodulators ( Figure 1 ) . This neuromodulator broadcasts information about inconsistency between the reward and the value function encoded by the population of critic neurons to all neurons in the network . The second term is synapse-specific and reflects the coincidence of EPSPs caused by presynaptic spikes of neuron with the postsynaptic spikes of neuron . The postsynaptic term is a consequence of the exponential non-linearity used in the neuron model ( see Models ) . This coincidence , “Hebbian” term is in turn filtered through the kernel which corresponds to the effect of a postsynaptic spike on . It reflects the responsibility of the synapse in the recent value function . Together these two terms form a three-factor rule , where the pre- and postsynaptic activities combine with the global signal to modify synaptic strengths ( Figure 2A , top ) . Because it has , roughly , the form of “TD error signalHebbian LTP” , we call this learning rule TD-LTP . We would like to point out the similarity of the TD-LTP learning rule to a reward-modulated spike-timing-dependent plasticity rule we call R-STDP [6] , [16] , [30]–[32] . In R-STDP , the effects of classic STDP [33]–[36] are stored into an exponentially decaying , medium term ( time constant ) , synapse-specific memory , called an eligibility trace . This trace is only imprinted into the actual synaptic weights when a global , neuromodulatory success signal is sent to the synapses . In R-STDP , the neuromodulatory signal is the reward minus a baseline , i . e . , . It was shown [32] that for R-STDP to maximize reward , the baseline must precisely match the mean ( or expected ) reward . In this sense , is a reward prediction error signal; a system to compute this signal is needed . Since the TD error is also a reward prediction error signal , it seems natural to use instead of . This turns the reward-modulated learning rule R-STDP into a TD error-modulated TD-STDP rule ( Figure 2A , bottom ) . In this form , TD-STDP is very similar to TD-LTP . The major difference between the two is the influence of post-before-pre spike pairings on the learning rule: while these are ignored in TD-LTP , they cause a negative contribution to the coincidence detection in TD-STDP . The filtering kernel , which was introduced to filter the spike trains into differentiable firing rates serves a role similar to the eligibility trace in R-STDP , and also in the discrete TD ( ) [1] . As noted in the previous section , this is the consequence of the combination of a smooth parametric function approximation of the value function ( each critic spike contributes a shape to ) and the form of the learning rule from Eq . 10 . The filtering kernel is crucial to back-propagation of the TD error , and thus to the solving of the temporal credit assignment problem . Having shown how spiking neurons can represent and learn the value function , we next test these results through simulations . However , in the actor-critic framework , the actor and the critic learn in collaboration , making it hard to disentangle the effects of learning in either of the two . To isolate learning by the critic and disregard potential problems of the actor , we temporarily sidestep this difficulty by using a forced action setup . We transform the water-maze into a linear track , and “clamp” the action choice to a value which leads the agent straight to the reward . In other words , the actor neurons are not simulated , see Figure 2B , and the agent simply “runs” to the goal . Upon reaching it at time , a reward is delivered and the trial ends . Figure 2C shows the value function over color-coded trials ( from blue to red ) as learned by a critic using the learning rule we described above . On the first run ( dark blue trace ) , the critic neurons are naive about the reward and therefore represent a ( noisy version of a ) zero value function . Upon reaching the goal , the TD error ( Figure 2D ) matches the reward time course , . According to the learning rule in Eq . 17 , this causes strengthening of those synapses that underwent pre-post activity recently before the reward ( with “recent” defined by the kernel ) . This is visible already at the second trial , when the value just before reward becomes positive . In the next trials , this effect repeats , until the TD error vanishes . Suppose that , in a specific trials , reward starts at the time when the agent has reached the goal . According to the definition of the TD error , for all times the -value is self consistent only if — or equivalently . The gray dashed line in Figure 2C shows the time course of the theoretical value function; over many repetitions the colored traces , representing the value function in the different trials , move closer and closer to the theoretical value . The black line in Figure 2C represents the average value function over 20 late trials , after learning has converged: it nicely matches the theoretical value . An interesting point that appears in Figure 2C is the clearly visible back-propagation of information about the reward expressed in the shape of the value function . In the first trials , the value function rises only for a short time just prior to the reward time . This causes , in the following trial , a TD error at earlier times . As trials proceed , synaptic weights corresponding to even earlier times increase . After trials in Figure 2C , the value function roughly matches the theoretical value just prior to , but not earlier . In subsequent trials , the point of mismatch is pushed back in time . This back-propagation phenomenon is a signature of TD learning algorithms . Two things should be noted here . Firstly , the speed with which the back-propagation occurs is governed by the shape of the kernel in the Hebbian part of the learning rule . It plays a role equivalent to the eligibility trace in reinforcement learning: it “flags” a synapse after it underwent pre-before-post activity with a decaying trace , a trace that is only consolidated into a weight change when a global confirmation signal arrives . This “eligibility trace” role of is distinct from its original role in the term , where it is used to smooth the spiking activity of the critic neurons ( Eq . 12 ) . As such , one might be tempted to change the decay time constant of the term in the learning rule so as to control back-propagation speed , while keeping the “other” of the signal fixed . In separate simulations ( not shown ) , we found that such an ad-hoc approach did not lead to a gain in learning performance . Secondly , we know by construction that this back-propagation of the reward information is driven by the TD error signal . However , visual inspection of Figure 2D , which shows the traces corresponding to the experiment in Figure 2C , does not reveal any clear back-propagation of the TD error . For , a large peak mirroring the reward signal ( gray dashed line ) is visible in the early traces ( blue lines ) and recedes quickly as the value function correctly learns to expect the reward . For , the is dominated by fast noise , masking any back-propagation of the error signal , even though the fact that the value function is learned properly shows it is indeed present and effective . One might speculate that if a biological system was using such a TD error learning system with spiking neuron , and if an experimenter was to record a handful of critic neurons he would be at great pain to measure any significant TD error back-propagation . This is a possible explanation for the fact that no back-propagation signal has been observed in experiments . We have already discussed the structural similarity of a TD-modulated version of the R-STDP rule [6] , [30] , [31] with TD-LTP . Simulations of the linear track experiment with the TD-STDP rule show that it behaves similarly to our learning rule ( data not shown ) , i . e . , the difference between the two rules ( the post-before-pre part of the coincidence detection window , see Figure 2A ) does not appear to play a crucial role in this case . We have seen above that spiking neurons in the “critic” population can learn to represent the expected rewards . We next ask how a spiking neuron agent chooses its actions so as to maximize the reward . In the classical description of reinforcement learning , actions , like states and time , are discrete . While discrete actions can occur , for example when a laboratory animal has to choose which lever to press , most motor actions , such as hand reaching or locomotion in space , are more naturally described by continuous variables . Even though an animal only has a finite number of neurons , neural coding schemes such as population vector coding [37] allow a discrete number of neurons to code for a continuum of actions . We follow the population coding approach and define the actor as a group of spiking neurons ( Figure 3A ) , each coding for a different direction of motion . Like the critic neurons , these actor neurons receive connections from place cells , representing the current position of the agent . The spike trains generated by these neurons are filtered to produce a smooth firing rate , which is then multiplied by each neuron's preferred direction ( see Models for all calculation details ) . We finally sum these vectors to obtain the actual agent action at that particular time . To ensure a clear choice of actions , we use a -winner-take-all lateral connectivity scheme: each neuron excites the neurons with similar tuning and inhibits all other neurons ( Figure 3B ) . We manually adjusted the connection strength so that there was always a single “bump” of neurons active . An example of the activity in the pool of actor neurons and the corresponding action readout over a ( successful ) trial is given in Figure 3C . The corresponding maze trajectory is shown in Figure 3D . In reinforcement learning , a successful agent has to balance exploration of unvisited states and actions in the search for new rewards , and exploitation of previously successful strategies . In our network , the exploration/exploitation balance is the result of the bump dynamics . To see this , let us consider a naive agent , characterized by uniform connections from the place cells to the actor neurons . For this agent , the bump first forms at random and then drifts without preference in the action space . This corresponds to random action choices , or full exploration . After the agent has been rewarded for reaching the goal , synaptic weights linking particular place cells to a particular action will be strengthened . This will increase the probability that the bump forms for that action the next time over . Thus the action choice will become more deterministic , and the agent will exploit the knowledge it has acquired over previous trials . Here , we propose to use the same learning rule for the actor neurons' synapses as for those of the critic neurons . The reason is the following . Let us look at the case where : the critic is signaling that the recent sequence of actions taken by the agent has caused an unexpected reward . This means that the association between the action neurons that have recently been active and the state neurons whose input they have received should be strengthened so that the same action is more likely to be taken again in the next occurrence of that state . In the contrary case of a negative reinforcement signal , the connectivity to recently active action neurons should be weakened so that recently taken action are less likely to be taken again , leaving the way to , hopefully , better alternatives . This is similar to the way in which the synapses from the state input to the critic neurons should be strengthened or weakened , depending on their pre- and postsynaptic activities . This suggests that the action neurons should use the same synaptic learning rule as the one in Eq . 17 , with now denoting the activity of the action neurons , but the signal still driven by the critic activity . This is biologically plausible and consistent with our assumption that is communicated by a neuromodulator , which broadcasts information over a large fraction of the brain . There are two critical effects of our -winner-take-all lateral connectivity scheme . Firstly , it ensures that only neurons coding for similar actions can be active at the same time . Because of the Hebbian part of the learning rule , this means that only those which are directly responsible for the action choice are subject to reinforcement , positive or negative . Secondly , by forcing the activity of the action neurons to take the shape of a group of similarly tuned neurons , it effectively causes generalization across actions: neurons coding for actions similar to the one chosen will also be active , and thus will also be given credit for the outcome of the action [16] . This is similar to the way the actor learns in non-neural actor-critic algorithms [18] , [19] , where only actions actually taken are credited by the learning rule . Thus , although an infinite number of actions are possible at each position , the agent does not have to explore every single one of them ( an infinitely long task ! ) to learn the right strategy . The fact that both the actor and the critic use the same learning rule is in contrast with the original formulation of the actor-critic network of Barto et al . [18] , where the critic learning rule is of the form “TD error×presynaptic activity” . As discussed above , the “TD error×Hebbian LTP” form of the critic learning rule Eq . 17 used here is a result of the exponential non-linearity used in the neuron model . Using the same learning rule for the critic and the actor has the interesting property that a single biological plasticity mechanism has to be postulated to explain learning in both structures . In the Morris water-maze , a rat or a mouse swims in an opaque-water pool , in search of a submerged platform . It is assumed that the animal is mildly inconvenienced by the water , and is actively seeking refuge on the platform , the reaching of which it experiences as a positive ( rewarding ) event . In our simulated navigation task , the learning agent ( modeling the animal ) is randomly placed at one out of four possible starting locations and moves in the two-dimensional space representing the pool ( Figure 4A ) . Its goal is to reach the goal area ( of the total area ) which triggers the delivery of a reward signal and the end of the trial . Because the attractor dynamics in the pool of actor neurons make it natural for the agent to follow a straight line , we made the problem harder by surrounding the goal with a U-shaped obstacle so that from three out of four starting positions , the agent has to turn at least once to reach the target . Obstacles in the maze cause punishment ( negative reward ) when touched . Similar to what is customary in animal experiments , unsuccessful trials were interrupted ( without reward delivery ) when they exceeded a maximum duration . During a trial , the synapses continually update their efficacies according to the learning rule , Eq . 17 . When a trial ends , we simulate the animal being picked up from the pool by suppressing all place cell activity . This results in a quick fading away of all neural activity , causing the filtered Hebbian term in the learning rule to vanish and learning to effectively stop . After an inter-trial interval of 3s , the agent was positioned in a new random position , starting a new trial . Figure 4B shows color-coded trajectories for a typical simulated agent . The naive agent spends most of the early trials ( blue traces ) learning to avoid walls and obstacles . The agent then encounters the goal , first at random through exploration , then repeatedly through reinforcement of the successful trajectories . Later trials ( yellow to red traces ) show that the agent mostly exploits stereotypical trajectories it has learned to reach the target . We can get interesting insight into what was learned during the trials shown in Figure 4B by examining the weight of the synapses from the place cells to actor or critic neurons . Figure 4C shows the input strength to critic neurons as a color map for every possible position of the agent . This is in effect a “value map”: the value the agent attributes to each position in the maze . In the same graph , the synaptic weights to the actor neurons are illustrated by a vector field representing a “policy preference map” . It is only a preference map , not a real policy map because the input from the place cells ( represented by the arrows ) compete with the lateral dynamics of the actor network , which is history-dependent ( not represented ) . The value and policy maps that were learned are experience-dependent and unique to each agent: the agent shown in Figure 4B and C first discovered how to reach the target from the “north” ( N ) starting position . It then discovered how to get to the N position from starting positions E and W , and finally to get to W from S . It has not however discovered the way from S to E . For that reason the value it attributes to the SE quarter is lower than to the symmetrically equivalent quarter SW . Similarly the policy in the SE quarter is essentially undefined , whereas the policy in the SW quarter clearly points in the correct direction . Figure 4D shows the distribution of latency – the time it takes to reach the goal – as a function of trials , for 100 agents . Trials of naive agents end after an average of ( trials were interrupted after ) . This value quickly decreases for agents using the TD-LTP learning rule ( green ) , as they learn to reach the reward reliably in about trials . We previously remarked that the TD-LTP rule of Eq . 17 is similar to TD-STDP , the TD-modulated version of the R-STDP rule [6] , [30] , [31] , at least in form . To see whether they are also similar in effect , in our context , we simulated agents using the TD-STDP learning rule ( for both critic and actor synapses ) . The blue line in Figure 4D show that the performance was only slightly worse than that of the TD-LTP rule , confirming our finding on the linear track that both rules are functionally equivalent . Policy gradient methods [5] follow a very different approach to reinforcement learning to TD methods . A policy gradient method for spiking neurons is R-max [4] , [6] , [32] , [38] , [39] . In short , R-max works by calculating the covariance between Hebbian pre-before-post activity and reward . Because this calculation relies on averaging those values over many trials , R-max is an inherently slow rule , typically learning on hundreds or thousands of trials . One would therefore expect that it can't match the speed of learning of TD-LTP or TD-STDP . Another difference of R-max with the other learning rules studied is that it does not need a critic [32] . Therefore we simulated an agent using R-max that only had an actor , and replaced the TD-signal by the reward , . The red line of Figure 4 show that , as expected , R-max agents learn much slower than previously simulated agent , if at all: learning is actually so slow , consistent with the usual timescales for that learning rule , that it can't be seen in the graph because this would require much longer simulations . One might object that using the R-max rule without a critic is unfair , and that it might benefit from a translation into a R-max rule with R = TD , by replacing the reward term by the error , as we did for R-STDP . But this overlooks two points . Firstly , such a “TD-max” rule could not be used to learn the critic: by construction , it would tend to maximize the TD error , which is the opposite of what the critic has to achieve . Secondly , even if one were to use a different rule ( e . g . TD-LTP ) to learn the critic , this would not solve the slow timescale problem . We experimented with agents using a “TD-max” actor while keeping TD-LTP for the critic , but could not find notable improvement over agents with an R-max actor ( data not shown ) . Having shown that our actor-critic system could learn a navigation task , we now address a task that requires higher temporal accuracy and higher dimensional state spaces . We focus on the acrobot swing-up task , a standard control task in the reinforcement control literature . Here , the goal is to lift the outermost tip of a double pendulum under the influence of gravity above a certain level , using only a weak torque at the joint ( Figure 5A ) . The problem is similar to that of a gymnast hanging below an horizontal bar: her hands rotate freely around the bar , and the only way to induce motion is by twist of her hips . While a strong athlete might be able to lift her legs above her head in a single motion , our acrobot is too weak to manage this . Instead , the successful strategy consists in moving the legs back and forth to start a swinging motion , building up energy , until the legs reach the sufficient height . The position of the acrobot is fully described by two angles , and ( see Figure 5A ) . However , the swinging motion required to solve the task means that even in the same angular position , different actions ( torque ) might be required , depending on whether the system is currently swinging to the left or to the right . For this reason , the angular velocities and are also important variables . Together , these four variables represent the state of the agent , the four-dimensional equivalent of the x–y coordinates in the navigation task . Just as in the water-maze case , place cells firing rates were tuned to specific points in the 4-dimensional space . Again similar to the maze navigation , the choice of the action ( in this case the torque exerted on the pendulum joint ) is encoded by the population vector of the actor neurons . The only two differences to the actor in the water-maze are that ( i ) the action is described by a single scalar and ( ii ) the action neuron attractor network is not on a closed ring anymore , but rather an open segment , encoding torques in the range . Several factors make the acrobot task harder than the water-maze navigation task . First , the state space is larger , with four dimensions against two . Because the number of place cells we use to represent the state of the agent grows exponentially with the dimension of the state space , this is a critical point . A larger number of place cells means that each is visited less often by the agent , making learning slower . At even higher dimensions , at some point the place cells approach is expected to fail . However , we want to show that it can still succeed in four dimensions . A second difficulty arises from the faster dynamics of the acrobot system with respect to the neural network dynamics . Although in simulations we are in full control of the timescales of both the navigation and acrobot dynamics , we wish to keep them in range with what might naturally occur for animals . As such the acrobot model requires fast control , with precision on the order of 100ms . Finally , the acrobot exhibits complex dynamics , chaotic in the control-less case . Whereas the optimal strategy for the navigation task consists in choosing an action ( i . e . , a direction ) and sticking to it , solving the acrobot task requires precisely timed actions to successfully swing the pendulum out of its gravity well . In spite of these difficulties , our actor-critic network using the TD-LTP learning rule is able to solve the acrobot task , as Figure 5B shows . We compared the performance to a near-optimal trajectory [40]: although our agents are typically twice as slow to reach the goal , they still learn reasonable solutions to the problem . Because the agents start with mildly random initial synaptic weights ( see Models ) and are subject to stochasticity , their history , and thus their performance , vary; the best agents have performance approaching that of the optimal controller ( blue trace in Figure 5B ) . We next try our spiking neuron actor-critic network on a harder control task , the cartpole swing-up problem [19] . This is a more difficult extension of cartpole balancing , a standard task in machine learning [18] , [41] . Here , a pole is attached to a wheeled cart , itself free to move on a rail of limited length . The pole can swing freely around its axle ( it doesn't collide with the rail ) . The goal is to swing the pole upright , and , ideally , to keep it in that position for as long as possible . The only control that can be exerted on the system is a force on the cart ( Figure 6A ) . As in the acrobot task , four variables are needed to describe the system: the position of the cart , its velocity , and the angle and angular velocity of the pole . We define a successful trial as a trial where the pole was kept upright ( ) for more than 10 s , out of a maximum trial length of . A trial is interrupted and the agent is punished for either hitting the edges of the rail ( ) or “over-rotating” ( ) . Agents are rewarded ( or punished ) with a reward rate . The cartpole task is significantly harder than the acrobot task and the navigation task . In the two latter ones , the agent only has to reach a certain region of the state space ( the platform in the maze , or a certain height for the acrobot ) to be rewarded and to cause the end of the trial . In contrast , the agent controlling the cartpole system must reach the region of the state space corresponding to the pole being upright ( an unstable manifold ) , and must learn to fight adverse dynamics to stay in that position . For this reason learning to successfully control the cartpole system takes a large number of trials . In Figure 6B , we show the number of successful trials as a function of trial number . It takes the “median agent” ( black line ) on the order of 3500 trials to achieve 100 successful trials . This is slightly worse but on the same order of magnitude as the ( non-neural ) actor-critic of [19] , which needs trials to reach that performance . The evolution of average reward by trial ( Figure 6C ) shows that agents start with a phase of relatively quick progression ( inset ) , corresponding to the agents learning to avoid the immediate hazard of running into the edges of the rail . This is followed by slower learning , as the agents learn to swing and control the pole better and better . To ease the long learning process we resorted to variable learning rates for both the actor and critic on the cartpole task: we used the average recent rewards obtained to choose the learning rate ( see Models ) . More precisely , when the reward was low , agents used a large learning rate , but when performance improved , the agents were able to learn finer control strategies with a small learning rate . Eventually agents manage fine control and easily recover from unstable situations ( Figure 6D ) . Detailed analysis of the simulation results showed that our learning agents suffered from noise in the actor part of the network , hampering the fine control needed to keep the pole upright . For example , the agent in Figure 6D has learned how to recover from a falling pole ( top and middle plots ) but will occasionally take more time than strictly necessary to bring the pole to a vertical standstill ( bottom plot ) . The additional spike firing noise in our spiking neuron implementation could potentially explain the performance difference with the actor-critic in [19] .
Throughout this study we tried to keep a balance between model simplicity and biological plausibility . Our network model is meant to be as simple and general as possible for an actor-critic architecture . We don't want to map it to a particular brain structure , but candidate mappings have already been proposed [42] , [43] . Although they do not describe particular brain areas , most components of our network resemble brain structures . Our place cells are very close to – and indeed inspired by – hippocampal place cells [22] . Here we assume that the information encoded in place cells is available to the rest of the brain . Actor neurons , tuned to a particular action and linked to the animal level action through population vector coding are similar to classical models of motor or pre-motor cortices [37] . So-called “ramp” neurons of the ventral striatum have long been regarded as plausible candidates for critic neurons: their ramp activity in the approach of rewards matches that of the theoretical critic . If one compares experimental data ( for example Figure 7A , adapted from van der Meer and Redish [44] ) and the activity of a typical critic neuron ( Figure 7B ) , the resemblance is striking . The prime neuromodulatory candidate to transmit the global TD error signal to the synapses is dopamine: dopaminergic neurons have long been known to exhibit TD-like activity patterns [7] , [45] . A problem of representing the TD error by dopamine concentration is that while the theoretically defined error signal can be positive as well as negative , dopamine concentration values [DA] are naturally bound to positive values [46] . This could be circumvented by positing a non-linear relation between the two values ( e . g . , ) at the price of sensitivity changes over the range . Even a simpler , piecewise linear scheme ( where is the baseline dopamine concentration ) would be sufficient , because learning works as long as the sign of the TD error is correct . Another possibility would be for the TD error to be carried in the positive range by dopamine , and in the negative range by some other neuromodulator . Serotonin , which appears to play a role similar to negative TD errors in reversal learning [47] , is a candidate . On the other hand this role of serotonin is seriously challenged by experimental recordings of the activity of dorsal raphe serotonin neurons during learning tasks [48] , [49] , which fail to show activity patterns corresponding to an inverse TD signal . One of the aspects of our actor-critic model that was not implemented directly by spiking neurons but algorithmically , is the computation of the TD signal which depends on the reward , the value function and its derivative . In our model , this computation is crucial to the functioning of the whole . Addition and subtraction of the reward and the value function could be done through concurrent excitatory and inhibitory input onto a group of neurons . Similarly , the derivative of the value function could be done by direct excitation by a signal and delayed ( for example by a an extra synapse ) inhibition by the same signal ( see example in Figure 7C ) . It remains to be seen whether such a circuit can effectively be used to compute a useful TD error . At any rate , connections from the the ventral striatum ( putative critic ) to the substantia nigra pars compacta ( putative TD signal sender ) show many excitatory and inhibitory pathways , in particular through the globus pallidus , which could have precisely this function [50] . A crucial limitation of our approach is that we rely on relatively low-dimensional state and action representations . Because both use similar tuning/place cells representations , the number of neurons to represent these spaces has to grow exponentially with the number of dimensions , an example of the curse of dimensionality . While we show that we can still successfully solve problems with four-dimensional state description , this approach is bound to fail sooner or later , as dimensionality increases . Instead , the solution probably lies in “smart” pre-processing of the state space , to delineate useful and reward-relevant low dimensional manifolds on which place cells could be tuned . Indeed , the representation by place cells can be learned from visual input with thousands of “retinal” pixels , using standard unsupervised Hebbian learning rules [20] , [51] , [52] . Moreover , TD-LTP is derived with the assumption of sparse neural coding , with neurons having narrow tuning curves . This is in contrast to covariance-based learning rules [53] , such as R-max [4] , [6] , [38] , [39] which can , in theory , work with any coding scheme , albeit at the price of learning orders of magnitude slower . Although a number of experimental studies exist [11]–[14] , [54] targeting the relation between STDP and dopamine neuromodulation , one is at pain to draw precise conclusions as to how these two mechanism interplay in the brain . As such , it is hard to extract a precise learning rule from the experimental data . On the other hand , we can examine our TD-LTP learning rule in the light of experimental findings and see whether they match , i . e . , whether a biological synapse implementing TD-LTP would produce the observed results . Experiments combining various forms of dopamine or dopamine receptor manipulation with high-frequency stimulation protocols at the cortico-striatal synapses provide evidence of an interaction between dopamine and synaptic plasticity [8]–[11] . While these experiments are too coarse to resolve the spike-timing dependence , they form a picture of the dopamine dependence: it appears that at high concentration the effect of dopamine paired with high-frequency stimulation is the induction of long-term potentiation ( LTP ) , while at lower concentrations , long-term depression ( LTD ) is observed . At a middle “baseline” concentration , no change is observed . This picture is consistent with TD-LTP or TD-STDP if one assumes a relation between the dopamine concentration and the TD error . The major difference between TD-LTP and TD-STDP is the behavior of the rule on post-before-pre spike pairings . While TD-LTP ignores these , TD-STDP causes LTD ( resp . LTP ) for positive ( resp . negative ) neuromodulation . Importantly this doesn't seem to play a large role for the learning capability of the rule , i . e . , the pre-before-post is the only crucial part . This is interesting in the light of the study by Zhang et al . [13] on hippocampal synapses , that finds that extracellular dopamine puffs reverse the post-before-pre side of the learning window , while strengthening the pre-before-post side . This is compatible with the fact that polarity of the post-before-pre side of the learning window is not crucial to reward-based learning , and might serve another function . One result predicted by both TD-LTP and TD-STDP and that has not , to our knowledge , been observed experimentally , is the sign reversal of the pre-before-post under negative reward-prediction-error signals . This could be a result of the experimental challenges required to lower dopamine concentrations without reaching pathological levels of dopamine depression . However high-frequency stimulation-based experiments show that a reversal of the global polarity of long-term plasticity indeed happens [8] , [11] . Moreover , a study by Seol et al . [54] of STDP induction protocols under different ( unfortunately not dopaminergic ) neuromodulators shows that both sides of the STDP learning window can be altered in both polarity and strength . This shows that a sign change of the effect of the pre-then-post spike-pairings is at least within reach of the synaptic molecular machinery . Another prediction that stems from the present work is the existence of eligibility traces , closing the temporal gap between the fast time requirements of STDP and delayed rewards . The concept of eligibility traces is well explored in reinforcement learning [1] , [5] , [55] , [56] , and has previously been proposed for reward-modulated STDP rules [6] , [30] . Although our derivation of TD-LTP reaches an eligibility trace by a different path ( filtering of the spike train signal , rather than explicitly solving the temporal credit assignment problem ) , the result is functionally the same . In particular , the time scales of the eligibility traces we propose , on the order of hundreds of milliseconds , are of the same magnitude as those proposed in models of reward-modulated STDP [6] , [30] . Direct experimental evidence of eligibility traces still lacks , but they are reminiscent of the synaptic tagging mechanism [57] . Mathematical models of tagging [58] , using molecular cascades with varying timescales , provide an example of how eligibility traces could be implemented physiologically . One interesting result of our study , is the fact that although our TD signal properly “teaches” the critic neurons the value function and back-propagates the reward information to more distant points , it is difficult to see the back-propagation in the time course of the TD signal itself . The reason for this is that the signal is drowned in rapid fluctuations . If one were to record a single neuron representing the TD error , it would probably be impossible to reconstruct the noiseless signal , except with an extremely high number of repetitions under the same conditions . This might be an explanation for the fact that the studies by Schultz and colleagues ( e . g . , [45] ) repeatedly fail to show back-propagation of the TD error , even though dopamine neurons seem to encode such a signal . In this study , TD-STDP ( and TD-LTP ) is used in a “gated-Hebbian” way: a synapse between A and B should be potentiated if it witnessed pre-before-post pairings and the TD signal following later is positive . This is fundamentally different from the role of the reward-modulated version of that learning rule ( R-STDP ) in [32] , where it is used to do covariance-based learning: a synapse between A and B should be potentiated if it witnesses positive correlation between pre-before-post pairings and a success signal , on average . One consequence of this is the timescale of learning: while TD-based learning takes tens of trials , covariance based learning typically requires hundreds or thousands of trials . The other side of the coin is that covariance-based learning is independent of the neural coding scheme , while TD-based learning requires neural tuning curves to map the relevant features prior to learning . The fact that the mathematical structure of the learning rule ( i . e . , a three-factor rule where the third factor “modulates” the effect of pre-post coincidences [59] ) is the same in both cases is remarkable , and one can see the advantage that the brain might have had to evolve such a multifunctional tool — a sort of “Swiss army knife” of synaptic plasticity .
For the actor and critic neurons we simulated a simplified spike response model ( , [60] ) . This model is a stochastic variant of the leaky integrate-and-fire neuron , with the membrane potential of neuron of given by ( 18 ) where is the efficacy of the synapse from neuron to neuron , is the set of firing times of neuron , is the membrane time constant , scales the refractory effect , is the Heaviside step function and is the last spike of neuron prior to . The EPSP is described by the time course ( 19 ) where is the synaptic rise time and is a scaling constant , and is the membrane time constant , as in Eq . 18 . Given the membrane potential , spike firing in the is an inhomogeneous Poisson process: at every moment the neuron has a probability of emitting a spike , according to an instantaneous firing rate ( 20 ) where , and are constants consistent with experimental values [61] . In the limit , the becomes a deterministic leaky integrate-and-fire neuron . The Morris water-maze pool is modeled by a two-dimensional plane delimited by a square wall . The position of the agent on the plane obeys ( 21 ) When the agent is within boundaries it moves with speed , as defined by the actor neurons' activity ( Eq . 29 ) . Whenever the agent encounters a wall , it instantly “bounces” back a distance along unitary vector , which points inward , perpendicular to the obstacle surface . Every “bumping” against a wall is accompanied by a punishing , negative reward delivery ( see reward delivery dynamics below ) . We used two variants of the navigation task . The linear track is a narrow rectangle of size centered around the origin , featuring a single starting position in and a wide goal area ( ) on the opposite side . Because the goal of this setup is to study critic learning , the action is clamped to a fixed value , so that the agent runs toward the goal at a fixed speed . The second variant is the navigation maze with obstacle . It consists of a square area of size centered around the origin , with four starting positions at . The goal area is a circle of radius centered in the middle of the maze . The goal is surrounded on three sides by a U-shaped obstacle ( width of each segment: 2 , length: 10 ) . In both variants , place cells centers are disposed on a grid ( blue dots on Figure 1 ) , with spacing coinciding with the width of the place fields . The outermost centers lie a distance outside the maze boundaries . This ensures a smooth coverage of the whole state space . In the case of the maze , the place cell grid consists of centers . For the linear track setup , the grid has centers . Trials start with the agent's position being randomly chosen from one out of four possible starting positions . The place cells , indexed by , are inhomogeneous Poisson processes . After a trial starts , the place cells' instantaneous firing rates are updated to ( 22 ) where is a constant regulating the activity of the place cells , is the place cells separation distance and the are the place cells centers . The presynaptic activity in the place cells generates activity in the post-synaptic neurons of the critic and the actor with a small delay caused by the rise time of EPSPs . The value function is calculated according to Eqs 12 and 13 , with parameters and . Because is delayed by the rise time of the kernel , at the start of a trial the TD error is subject to large , boundary effect transients . To cancel these artifacts , we clamp the TD error to , for the first of each trial . We use a reward discount time constant . The goal of the agent is to reach the circular area which represents the submerged platform of the water-maze . When the agent reaches this platform , a positive reward is delivered , the trial ends and the agent is put in a so-called “neutral state” , which models the removal of the animal from the experiment area . The effects of this is ( i ) the place cells corresponding to the maze become silent , presumably replaced by other ( not modeled ) place cells , and ( ii ) the expectation of the animal becomes neutral , and therefore its value function goes to zero . So at the end of a trial , we turn off place cell activity ( ) , and the value function is no longer given by Eq . 13 , but decays exponentially to 0 with time constant from its value at the time of the end of the trial . Importantly , synaptic plasticity continues after the end of the trial , so that the effect of affects the synaptic weight even though its delivery takes place in the neutral state . Additionally , a trial can end without the platform being reached: if a trial exceeds the time limit , it is declared a failed trial , and interrupted with the agent put in the neutral state , just as in the successful case , but without reward being delivered . According to Eq . 3 , rewards are given to the agent as a reward rate . This reflects the fact that “natural” rewards , and reward consumption , are spread over time , rather than point-like events . So we translate absolute rewards ( ) to a reward rate ( ) , calculated as the difference of two decaying “traces” obeying dynamics ( 23 ) i . e . , ( 24 ) At most times , the reward is close to 0 . Reward is delivered only when some event ( goal reached or collision against an obstacle ) occurs . The delivery of a reward happens through instantaneous update of the traces ( 25 ) The resulting effect is a subsequent positive excursion of , with rise time and fall time , which , integrated over time , amounts to . In the acrobot task , the position of the pendulum is described by two angles: is the angle between the first segment of the pendulum and the vertical , and is the angle between the second segment and an imaginary prolongation of the first ( Figure 5A ) . When , the pendulum hangs down . Critical to solving the task are also the angular velocities and . As in the maze navigation case , place cells tuned to specific centers are used to represent the state of the acrobot . We transform the angular velocities , . This allows a fine resolution over small velocities , while maintaining a representation of higher velocities with a small number of place cells . The state is represented by the four variables . The place cells centers are disposed on a 4-dimensional grid defined by indexes , such that with ( 26 ) This yields a total of centers . The activity of a place cell with center is defined by ( 27 ) where is a function returning the difference between two angles modulo in the range and the place cell widths to correspond to the grid spacing as in Eq . 26 . The acrobot dynamics obeys the following equations [1]:Here , , , and are convenience variables , is the torque applied to the joint , are the lengths of the segments , of mass , with moments of inertia and lengths to the centers of mass , under the influence of gravity . All dimensions except time are unit-less . The goal is for the tip of the acrobot to reach a height over the axis , i . e . , fulfill the condition . Once this happens , or the maximum trial time is reached , the trial ends . To entice the acrobot to do something , we give an ongoing punishment to the agent for not reaching the reward , to be compared with the reward received at the goal . As in the water-maze case , we use a reward discount time constant . Due to the larger number of place cells , we use less critic and actor neurons than in the maze case , respectively and , to reduce the number of synapses and the computational load . To compare the performance of our agent against an “optimal” strategy , we use the direct search method [40] . The main idea behind the method is to search for the sequence of action that will maximize the system's total energy , with knowledge of the acrobot dynamics . To make the search computationally tractable , a few simplifications are made: actions are limited to the alternative , actions are only taken in steps of 100 ms , only a window of the next 10 steps is considered at a time , and the number of action switch in each window is limited to 2 . Thus only 55 action sequences have to be examined , and the sequence that maximizes the total energy reached over the window , or reaches the goal height the sooner , is selected . The first action in that sequence is chosen as the action for the next step and the whole procedure is repeated with the window shifted by one step . The goal height reaching latency found with this method was 7 . 66s ( red line in Figure 5B ) . The position of the cartpole system is described by the cart position , the cart velocity , the angle of the pole with the vertical ( corresponds to the pole pointing upwards ) and the angular velocity ; these form the state vector . Similar to the acrobot , the place cells for the cartpole problem are regularly disposed on a four-dimensional grid of cells . The location of a place cell with index is at location with ( 28 ) The activity of a place cell is defined in a way analog to Eq . 27 . The variance of the gaussian place fields is diagonal , where corresponds to the grid spacing in dimension . The dynamics of the cartpole are [62]:Here a = v is the acceleration of the cart , is half the pole's length , and are coefficients of friction of the cart on the track and of pole rotation respectively . The cart , with mass , and the pole , with mass , are subject to the acceleration of gravity . As in the acrobot case , all dimensions except time are unit-less . Following [19] , the agent is rewarded continuously depending on the current height of the pole with , and the reward discount time constant is set to . If the cart runs off its rail ( ) or over-rotates ( ) the trial is ended and a negative reward is given . A trial ends without reward after . When a new trial starts , the position of the system is initialized with a random and . In population vector coding , each actor neuron “votes” for its preferred action in the action space , by firing an action potential . An action vector is obtained by averaging the product of the instantaneous firing rate ( see Eq . 12 ) and the action vector of each neuron , i . e . ( 29 ) where is defined as ( 30 ) with filterwith and being filtering time constants . The term in Eq . 29 is a normalization term . In the case of the navigation task ( two-dimensional action ) , it is equal to the number of actor neurons , . In the cases of the acrobot and the cartpole task ( scalar action ) , . We enforce a N-winner-takes-all mechanism on the action neurons by imposing “lateral” connectivity between the action neurons: action neurons coding for similar actions excite each other , while they inhibit the neurons coding for dissimilar actions . The synaptic weight between two action neurons and is ( 32 ) where is a lateral connectivity function . This is zero for , peaks for and monotonously decreases towards 0 as and diverge . is a normalization constant . The parameters and regulating the recurrent connections were manually tuned: the lateral connectivity has to be strong enough so that there is always exactly one “bump” of similarly tuned neurons active whenever the action neurons receive some excitation from the place cells , but not so strong that it completely dominates the feed-forward input from the place cells . The preferred vectors of the action neurons and the function are dependent on the learning task . In the case of the maze navigation task , the preferred action vectors are where is a constant representing the agent velocity per rate unit and , for . The function was chosen as ( 33 ) with . In the case of the acrobot and cartpole tasks , the action vectors are . For the acrobot represents the maximum torque that the agent can exert and for the cartpole task is the maximum force on the cart . The lateral connectivity function in both cases was chosen as ( 34 ) with . Additionally , we algorithmically constrain the torque exerted by the agent to the domain . This models the limited strength of the agent's “muscles” . In R-STDP [6] , [30]–[32] , the effects of classic STDP are modulated by a neuromodulatory signal , where is a constant baseline . We transformed the reward-modulated R-STDP into the TD-modulated rule TD-STDP by replacing the with . The TD-STDP rule can be written as ( 35 ) where the STDP learning window isThe eligibility trace kernel is the result of an exponential decay , i . e . , , with time constant . The positive constants and govern the size of the pre-before-post and post-before-pre parts of the learning window respectively , and the time constants and determine their timing requirement . R-max [4] , [6] , [32] , [38] is a reward-modulated learning rule derived from policy gradient principles [5] . It can be written as ( 36 ) where is the instantaneous firing rate of neuron , as defined in Eq . 20 . Initial values of the synaptic weights to both critic and actor were randomly drawn from a normal distribution with mean and standard deviation . These values ensured an initial value function and reasonable action neuron activity prior to learning . For all learning rules , synaptic weights were algorithmically constrained to the range , to avoid negative or runaway weights . Learning rate values were manually adjusted ( one value for actor and another one for critic synapses ) to the value that yielded the best performance ( as measured by the number of trials completed in 2 . 000s of simulated time ) . These values for the navigation and acrobot tasks are printed in Table 1 . For the cartpole task , somewhat faster learning was achieved by using a variable learning rate ( 37 ) for the critic , where is a running average of past reward rates , computed by filtering with an exponential window with time constant 50s . The actor learning rate was . All simulations were ran using Euler's method with time-step , except for the acrobot and cartpole dynamics , simulated using 4th order Runge-Kutta with . In this section we calculate the term , needed to derive Eq . 17 . Using Eqs 12–13 , and focusing on the synaptic weight from to , we find ( 38 ) where we used the fact that ρi′ ( t ) is independent of for . The derivative of the spike train is ill-defined: in our stochastic neuron model , the spike train itself is independent of the synaptic weights . It is only the probability of the spike train actually being emitted by the neuron that depends on the weights . Therefore we replace with , the expected value of the spike train conditional on the input . This yields ( 39 ) where the sum is over all possible spike trains and is the probability density of the spike train being equal to . The probability density of that spike train , lasting from to , being produced by an SRM0 neuron receiving inputs is [38] ( 40 ) where is the membrane potential ( Eq . 18 ) and we have used Eq . 20 . Combining Eqs 39 and 40 yields ( 41 ) The integration reflects the fact that the probability of a spike being emitted by the neuron at time is dependent not only on recent presynaptic spikes , but also on the time of the last spike of neuron , which in turn depends on its whole history . It is not clear that , in our context , this history dependence is a desirable outcome . Two devices already take the spike train history into account . Firstly , the definition of the value function in the TD framework is conditional only on the current state , and not the long-term history . ( This stems from the Markov decision process at the root of TD . ) Secondly , the filtering of the spike train by already ensures that the short-term history is remembered , making the integral over the history redundant . For these reasons , we choose to neglect the neuron's history , and to perform the following substitution ( 42 ) i . e . , we take the last spike time of neuron as given , and we ask how does the mean spiking at time vary as a function of the synaptic weight . Therefore we have ( 43 ) where we have used the definition of the neuron's firing rate , Eq . 20 , and is the Dirac distribution . Using Eqs 18 yields ( 44 ) where is the spike train of neuron culled to times posterior to the spikes of neuron , i . e . , , with denoting the Heaviside step function . Wrapping up the steps from Eqs 38 and 42–44 , we finally have ( 45 ) In the Results section we derive a learning rule starting from Eq . 10 . We also suggest that starting from a gradient descent on the squared TD error ( Eq . 17 ) should yield a valid learning rule . Here we derive such a learning rule . Combining Eq . 10 , the definition of the TD error ( Eq . 5 ) and the result of the previous section ( Eq . 45 ) , we find ( 46 ) where is the spike train of presynaptic neuron . This learning rule has the same general form as the TD-LTP rule ( Eq . 17 ) : a “Hebbian” pre-before-post coincidence term is first temporally filtered , and then multiplied by the TD error with a term ( Figure 8A ) . The difference lies in the extra in the filter , which comes from a term . As Figure 8 suggests , the term largely dominates over . This is the consequence of our choice of a long discount time constant ( ) with a short ( ) kernel . Here we show , both analytically and in simulations , that the squared TD gradient learning rule of Eq . 46 suffers from a noise bias problem . This arises from the noise in the individual neurons estimating the value function , and is serious enough to prevent learning . To see this , we start by decomposing the spike train of a neuron into a mean and a noise term , i . e . ( 47 ) where we have defined , with the brackets denoting expectation , i . e . , averaging over all possible outcomes of critic neurons activity conditioned on the presynaptic neural activity . With this definition , we can rewrite Eq . 46 as ( 48 ) where the error has been spelled out explicitly ( Eqs 5 , 13 and 12 ) . Eq . 48 suggests that quadratic terms in the noise might play a role in the learning rule . Indeed , distributivity and use of the facts and for gives ( 49 ) Here we have defined the autocorrelation of the noise terms , as well as for brevity . The first term in the right-hand side of Eq . 49 is analog to Eq . 46 , with replacing , and replacing . In effect this is a “mean” version of the learning rule: this is what one would get by replacing the stochastic spiking neurons in the model by analog , noiseless units with a similar exponential activation function . The second term arises from the correlation of neuron noise in the TD term and the Hebbian component of the learning rule . This term is a function of the autocorrelation function of the postsynaptic neuron . This carries only indirect information about the postsynaptic firing ( and thus the current value function ) and no information about the reward . For this reason , we conjecture that this second element is a potentially problematic term , which we refer to as the “nuisance” term . This hypothesis is confirmed by linear track simulations using the learning rule Eq . 46 , shown in Figure 8B . These indicate that the learning rule is unable to learn the task , contrary to TD-LTP ( Figure 8C , same as Figure 2B ) . More precisely , the value functions learned by the squared TD gradient rule suffer from a negative “drag” term . We next try to identify this negative “drag” with the nuisance term . Although there's no closed form expression for , one can use the statistics of a Poisson process as a first order approximation . In that case ( is the Dirac distribution ) and Eq . 49 becomes ( 50 ) The last term on the right-hand side of Eq . 50 implies that , on average , each presynaptic spike in neuron causes the synaptic weight to depress by a fixed amount . This quantity increases with the variance of the noise process , in this case the inhomogeneous Poisson process that drives the neuron , and inversely to the number of critic neurons . The time course of the presynaptic spike effect is ruled by , which is plotted in the top panel of Figure 8D . The aggregate nuisance effect on of a single presynaptic spike is proportional to the integral of over time . In Figure 8E , we explore the dependence of the nuisance term on in numerical simulations . Eq . 50 suggests that the mean learning rule term should obey a relationship of the form ( 51 ) Here is the result of the “useful” part of the learning rule , and contains all the other dependencies of the nuisance term . We tested the dependency by simulating agents with variable numbers of critic neurons in a linear track scenario . The setup was similar to that of Figure 2 , except that the weights were frozen , i . e . , we collected the value of the learning rule at each time step , but we didn't actually update the weights . The mean learning rule outcome for 200s of simulations are plotted in Figure 8E as crosses , against the number of critic neurons . The black line shows a fit of the data by Eq . 51: both are in good agreement . From Eq . 50 , we see that the nuisance term also depends on the variance of the noise process . It is difficult to control the variance of our spiking neurons' noise process without also altering their firing rate and thus the result of the learning rule . To circumvent this difficulty , we turned to a rate model , where the single critic neuron's firing rate was ( 52 ) where is a constant , the place cells rates are defined in Eq . 22 and is a white noise process . Similar to the steps above , a gradient descent on yields a learning rule of the form ( 53 ) Due to the noise component in , the learning rule suffers from the same noise-driven nuisance as the spiking version . This depends on the noise's variance , so that the mean weight change obeys ( 54 ) where . In Figure 8F , we use the rate-based model and rule in the same “frozen weights” linear track scenario as in Figure 8E . This time we looked at how the mean weight change varied as a function of the noise variance . Again , the data is well matched by a fit with Eq . 54 ( black line ) , suggesting that the nuisance term behaves as expected . In the preceding section we found that a noise correlation nuisance in the squared TD gradient learning rule causes it to be ineffective . However , the same actually should apply to the TD-LTP rule . Indeed , if we repeat the steps above leading to Eq . 50 for the learning rule TD-LTP , we get ( 55 ) The only difference is the time course of the nuisance term , which is for the squared TD gradient rule versus for TD-LTP . Figure 8D shows a plot of both expressions: because the TD-LTP expression is much smaller , these are plotted on different axes . As noted before , the integral of the nuisance is proportional to these time courses ( shown on Figure 8D ) . The term for TD-LTP is more than three orders of magnitude smaller than that of the square TD gradient rule . In Figure 8G and H , we repeat the experiments of Figure 8E and F , respectively . These show that the TD-LTP learning rule also suffers from a nuisance term , but that it is orders of magnitude smaller than for the squared TD gradient rule . As shown by Figure 8C and in the Results section , this nuisance is not sufficient to prevent TD-LTP from properly learning the value function . In the Results section , we claim that -values based algorithms , such as Sarsa [24] and Q-Learning [23] are difficult to extend to continuous time in a neural network setting . Here we develop this argument . In the discrete Sarsa algorithm , the agent maintains an estimation of the state-action -values . For an agent following the policy , starting at time step in state and executing action , this is defined as the discounted sum over future rewards : ( 56 ) Here is a discount factor , and and represent the future states and actions visited by the agent under policy . To learn -values approximations to the real , Sarsa suggests the following update rule at time step : ( 57 ) where the TD error is defined as ( 58 ) If one were to propose a continuous time version of Sarsa , one would start by redefining the state-action value function to continuous time t , similar to the value function of Eq . 3 ( 59 ) Here now plays the role of the discount factor . As we did for Eq . 5 , we define the TD error on the -value by taking the derivative of Eq . 59 ( 60 ) To calculate the TD error , one therefore needs to combine the three terms in Eq . 60 . We assume the reward is given by the environment . Typically [16] , [20] , neural networks implementations of -values based reinforcement learning consist of a number “action cells” neurons , each tuned to a specific action and rate-coding for the state-action values ( 61 ) where is neuron 's firing rate . In that case , reading out the value is thus simply a matter of reading the activity of the neuron coding for the action selected at time . Reading out the temporal derivative is harder to do in that context , because the currently chosen action is evolving all the time . For small , we can approximate ( 62 ) where we also used Eq . 61 and identified the action neuron tuned to action . The difficulty that arises in evaluating Eq . 62 is the following . It requires a system that can keep track of the two recent actions and , identify the relevant neurons and , and calculate a difference of their firing rates . This is hard to envision in a biologically plausible setting . The use of an actor-critic architecture solves this problem by having a single population coding for the state-based value at all times .
|
As every dog owner knows , animals repeat behaviors that earn them rewards . But what is the brain machinery that underlies this reward-based learning ? Experimental research points to plasticity of the synaptic connections between neurons , with an important role played by the neuromodulator dopamine , but the exact way synaptic activity and neuromodulation interact during learning is not precisely understood . Here we propose a model explaining how reward signals might interplay with synaptic plasticity , and use the model to solve a simulated maze navigation task . Our model extends an idea from the theory of reinforcement learning: one group of neurons form an “actor , ” responsible for choosing the direction of motion of the animal . Another group of neurons , the “critic , ” whose role is to predict the rewards the actor will gain , uses the mismatch between actual and expected reward to teach the synapses feeding both groups . Our learning agent learns to reliably navigate its maze to find the reward . Remarkably , the synaptic learning rule that we derive from theoretical considerations is similar to previous rules based on experimental evidence .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Models"
] |
[
"computational",
"neuroscience",
"biology",
"neuroscience",
"learning",
"and",
"memory"
] |
2013
|
Reinforcement Learning Using a Continuous Time Actor-Critic Framework with Spiking Neurons
|
Chemical-genetic interactions–observed when the treatment of mutant cells with chemical compounds reveals unexpected phenotypes–contain rich functional information linking compounds to their cellular modes of action . To systematically identify these interactions , an array of mutants is challenged with a compound and monitored for fitness defects , generating a chemical-genetic interaction profile that provides a quantitative , unbiased description of the cellular function ( s ) perturbed by the compound . Genetic interactions , obtained from genome-wide double-mutant screens , provide a key for interpreting the functional information contained in chemical-genetic interaction profiles . Despite the utility of this approach , integrative analyses of genetic and chemical-genetic interaction networks have not been systematically evaluated . We developed a method , called CG-TARGET ( Chemical Genetic Translation via A Reference Genetic nETwork ) , that integrates large-scale chemical-genetic interaction screening data with a genetic interaction network to predict the biological processes perturbed by compounds . In a recent publication , we applied CG-TARGET to a screen of nearly 14 , 000 chemical compounds in Saccharomyces cerevisiae , integrating this dataset with the global S . cerevisiae genetic interaction network to prioritize over 1500 compounds with high-confidence biological process predictions for further study . We present here a formal description and rigorous benchmarking of the CG-TARGET method , showing that , compared to alternative enrichment-based approaches , it achieves similar or better accuracy while substantially improving the ability to control the false discovery rate of biological process predictions . Additional investigation of the compatibility of chemical-genetic and genetic interaction profiles revealed that one-third of observed chemical-genetic interactions contributed to the highest-confidence biological process predictions and that negative chemical-genetic interactions overwhelmingly formed the basis of these predictions . We also present experimental validations of CG-TARGET-predicted tubulin polymerization and cell cycle progression inhibitors . Our approach successfully demonstrates the use of genetic interaction networks in the high-throughput functional annotation of compounds to biological processes .
The discovery of chemical compounds with desirable and interesting biological activity advances our understanding of how compounds and biological systems interact . Chemical-genetic interaction profiling enables this discovery by measuring the response of defined gene mutants to chemical compounds [1–8] . Specifically , a chemical-genetic interaction profile refers to the set of gene mutations that confer sensitivity ( a negative chemical-genetic interaction ) or resistance ( a positive interaction ) to a compound and provides functional insights into the compound’s mode ( s ) of action . Recent advances in DNA sequencing technology have enabled dramatic increases in the throughput of chemical-genetic interaction screens ( into the range of thousands of compounds ) via multiplexed analysis of pooled mutant libraries [6 , 7 , 9] Similarly , genetic interactions identify pairs of gene mutations whose combined phenotypes are more or less severe than expected given the phenotypes of the individual mutants . In S . cerevisiae , the vast majority of all possible gene double-mutant pairs have been constructed and scored for fitness-based genetic interactions , yielding a global compendium of genome-wide genetic interaction profiles that quantitatively describe each gene’s function . Similarity between two genes’ genetic interaction profiles implies that these genes perform similar functions , enabling the functional annotation of uncharacterized genes and the construction of a global hierarchy of cellular function [5 , 10] . The global genetic interaction network in S . cerevisiae provides a resource for interpreting chemical-genetic interaction profiles across a broad range of cellular function , as the chemical-genetic interaction profile of a compound should resemble the genetic interaction profile of its cellular target or target processes [2 , 5] . Importantly , this approach to interpretation does not depend on reference chemical-genetic interaction profiles and thus enables the discovery of compounds with novel modes of action . Previous small and large-scale chemical-genetic interaction studies have employed various computational methods to provide more informative clustering of the resulting interaction matrices [3 , 11] and even predict perturbed protein complexes [12] or direct protein targets [13] . However , the integration of chemical-genetic and genetic interaction profiles has only been performed in the context of relatively small studies [2 , 5] . Here , we present the use of genetic interaction profiles to systematically interpret chemical-genetic interaction profiles on a large scale . To this end , we developed a computational method , called CG-TARGET ( Chemical Genetic Translation via A Reference Genetic nETwork ) , that integrates chemical-genetic and genetic interaction profiles to predict the biological processes perturbed by compounds . In a recent publication [14] , we applied this method to a chemical-genetic interaction screen of nearly 14 , 000 compounds in S . cerevisiae [14] , using profiles from the global yeast genetic interaction network [5 , 10] to interpret the chemical-genetic interaction profiles . Here , we show that CG-TARGET recapitulates known information for well-characterized compounds and showed a marked improvement in false discovery rate control compared to alternative , enrichment-based approaches . Additionally , we experimentally validated two different mode-of-action predictions , one in an in vitro system using mammalian proteins , confirming both the accuracy of the predictions and the potential to translate them across species . CG-TARGET is available , free for non-commercial use , at https://github . com/csbio/CG-TARGET .
We obtained chemical-genetic interaction profiles from a recent large-scale chemical-genetic interaction screen in S . cerevisiae [14] . Profiles were obtained in two batches , labeled “RIKEN” and “NCI/NIH/GSK” to reflect the compound libraries screened–for RIKEN , the RIKEN Natural Product Depository [15] , and for NCI/NIH/GSK , plated libraries from the NCI Open Chemical Repository , the NIH Clinical Collection , and the GlaxoSmithKline Published Kinase Inhibitor Set [16] . The RIKEN compounds were primarily natural products and derivatives–mostly uncharacterized–but also contained ~200 approved drugs and chemical probes from which we selected a well-characterized subset for benchmarking . The NCI/NIH/GSK compounds were more characterized , having been tested against the NCI-60 cancer cell line panel ( NCI collections ) , tested in clinical trials ( NIH Clinical Collection ) or designed to inhibit human kinases ( GSK ) –but their specific modes of action remained primarily uncharacterized . The final datasets consisted of interaction scores for 8418 RIKEN compounds and 3565 NCI/NIH/GSK compounds ( with 5724 and 2128 negative control conditions , respectively ) screened against a diagnostic set of ~300 haploid gene deletion mutants selected to optimally capture the information in the complete S . cerevisiae non-essential deletion collection [14 , 17] . Each profile contained z-scores that reflected the deviation of each strain’s observed abundance from expected abundance in the presence of a compound . Genetic interaction profiles were obtained from a recently assembled , genome-wide compendium of genetic interaction profiles in S . cerevisiae [5] . These profiles were generated through the systematic analysis of double mutant fitness and consist of epsilon scores that reflect the deviation of each double mutant’s observed fitness from that expected given the single mutant fitness values , assuming a multiplicative null model [18] . Profiles were filtered to the ~35% with the highest signal , and we mapped these 1505 high-signal “query” genes to Gene Ontology biological process terms [19 , 20] to define the bioprocess targets of compounds . ( see Materials and Methods ) . We developed CG-TARGET ( Chemical Genetic Translation via A Reference Genetic nETwork ) to predict the biological processes perturbed by compounds in our recently-generated dataset of ~12 , 000 chemical-genetic interaction profiles ( Fig 1 ) . CG-TARGET requires three input datasets: 1 ) chemical-genetic interaction profiles; 2 ) genetic interaction profiles; and 3 ) a mapping from the query genes in the genetic interaction profiles to gene sets representing coherent biological processes ( referred to as “bioprocesses” ) . Predicting the bioprocesses perturbed by a particular compound involves four distinct steps . First , a control set of resampled chemical-genetic interaction profiles is generated , each of which consists of one randomly-sampled interaction score per gene mutant across all compound treatment profiles in the chemical-genetic interaction dataset; these profiles thus provide a means to account for variance in each mutant strain observed upon treatment with bioactive compound but not upon treatment with experimental controls ( DMSO with no active compound ) . Second , “gene-target” prediction scores between each compound and query gene are generated by computing an inner product between all chemical-genetic interaction profiles ( comprising compound treatment , experimental control , and random profiles ) and all L2-normalized query genetic interaction profiles; normalizing only the genetic interaction profiles results in gene-target scores that should be more robust to noise in the chemical-genetic data [21] and reflect the overall strength of each chemical-genetic profile as well as its similarity to gene mutants’ profiles . Third , these “gene-target” prediction scores are aggregated into bioprocess predictions; a z-score and empirical p-value for each compound-bioprocess prediction are obtained by mapping the gene-target prediction scores to the genes in the bioprocess of interest and comparing these scores to those from shuffled gene-target prediction scores and to distributions of the scores derived from experimental control and resampled profiles . Finally , the false discovery rates for these predictions are estimated by comparing , across a range of significance thresholds , the frequency at which experimental control and randomly resampled profiles predict bioprocesses versus that of compound treatment profiles ( see Materials and Methods ) . A schematic representation of the method is provided as S1 Fig . To provide a baseline for benchmarking the performance of CG-TARGET on these large screens , we implemented two simple , enrichment-based approaches for predicting bioprocess-level targets . The “direct enrichment” approach tested for enrichment of GO biological processes among each compound’s 20 strongest negative chemical-genetic interactors , providing a comparison to methods that do not incorporate genetic interaction profiles . The “gene-target enrichment” approach tested for the enrichment of GO biological processes among the top-n gene-target prediction scores for each compound , enabling a comparison of CG-TARGET’s z-score-based approach to enrichment on the gene-target scores . For the comparisons to gene-target enrichment below , we selected n = 20 as it showed the best overall performance across a range of values of n ( S2 Fig ) . We applied CG-TARGET to the RIKEN and NCI/NIH/GSK chemical-genetic interaction screens , identifying 848 out of 8418 compounds ( 10% ) from the RIKEN screen and 705 of 3565 compounds ( 20% ) from the NCI/NIH/GSK screen with at least one prediction that achieved false discovery rates of 25 and 27% , respectively ( referred to as “high-confidence” compounds and predictions ) ( Table 1 , Fig 2 ) . Measured using the RIKEN dataset , this rate of discovery at FDR ≤ 25% was over 4-fold higher in terms of number of discovered compounds than that of direct enrichment ( 190 compounds ) and over 100-fold higher than that of gene-target enrichment ( 7 compounds , Fig 3A ) . In all cases , the false discovery rates derived from resampled profiles were more conservative than those derived from experimental controls , suggesting that some sources of variance in each gene mutant’s interaction scores arose only upon treatment with compound and therefore could not be corrected using only solvent controls . In addition to assessing false discovery rate control relative to baseline methods , we also assessed prediction accuracy . We performed the first of these comparisons against the direct enrichment predictions by asking if the top prediction for each of 35 well-characterized compounds matched what was known about that compound . For direct enrichment , the top prediction for 11 of these 35 compounds matched its known mode of action , with only 6 of these compounds passing the FDR ≤ 25% criteria that would enable their discovery in a large-scale screen ( S1 Table ) . In contrast , CG-TARGET matched 17 of these compounds to their known mode of action , with 16 passing the FDR ≤ 25% discovery threshold . We then compared CG-TARGET to gene-target enrichment using two measures of accuracy . The first accuracy-based evaluation was performed on genetic interaction profiles with added noise , which provided a means to both simulate chemical-genetic interaction profiles and annotate them with gold-standard GO biological process annotations for evaluation . For the second accuracy-based evaluation , we assigned each of the aforementioned well-characterized compounds to a “gold standard” bioprocess term and evaluated the ranks of each compound’s gold-standard bioprocess within its list of bioprocess predictions . We note that neither of these methods were particularly suitable for comparing CG-TARGET to direct enrichment , as 1 ) the assumption of alignment between chemical-genetic and genetic interaction profiles was implicit in the generation of the simulated profiles and 2 ) we anticipated that spurious rank differences would result from differences in the size ( ~300 genes for direct versus ~1500 genes for CG-TARGET ) and composition ( about half of the former in the latter ) of the two gene universes that defined the bioprocess term sets . CG-TARGET performed comparably to the best-performing gene-target enrichment method using our measures of accuracy . This is first shown in the evaluation of these methods’ respective abilities to predict a gold-standard annotated bioprocess as the top prediction for each simulated chemical-genetic interaction profile . Specifically , CG-TARGET performed nearly as well as the top-20 gene-target enrichment method across both low and high recall values ( Fig 3B ) . Both methods captured a gold-standard annotation as the top predicted bioprocess for approximately 34% of the simulated compounds ( 33 . 4% and 35 . 6% for CG-TARGET and top-20 gene-target enrichment , respectively ) , which represented more than a 22-fold enrichment over the background expectation of 1 . 5% ( the average number of gold-standard bioprocess annotations per simulated compound divided by the number of bioprocesses ) . For the 35 gold-standard compound-bioprocess pairs , we observed that both CG-TARGET and gene-target enrichment captured the gold-standard bioprocess for 6 and 21 ( out of 35 ) compounds above ranks of 2 and 40 ( out of 1329 ) , respectively , with slightly decreased performance for CG-TARGET between these rank thresholds ( Fig 3C , Table 2 ) . The significance of these rank values was evaluated by randomizing the order of each compound’s bioprocess predictions 10 , 000 times and recalculating the ranks . Both methods achieved similar results in this respect , with CG-TARGET and gene-target enrichment respectively identifying 22 and 21 gold-standard compounds with significantly better ranks than the random expectation . The two methods also performed similarly when comparing the “effective rank” of each compound’s gold-standard bioprocess , with CG-TARGET and gene-target enrichment respectively identifying 20 and 22 compounds for which the gold-standard or a closely-related bioprocess achieved a rank of 5 or better . Despite the similar performance in rank space , however , none of the 21 significantly-ranked predictions made by gene-target enrichment achieved FDR ≤ 25% , compared to 16 out of 22 for CG-TARGET ( Table 2 ) . In addition to benchmarking CG-TARGET’s ability to prioritize gold-standard annotated bioprocesses for specific compounds , we also benchmarked its ability to prioritize compounds that perturb specific bioprocesses . Specifically , each GO term was evaluated based on the ranks of the predictions for the simulated chemical-genetic interaction profiles derived from genes annotated to that GO term . The 100 best-performing terms represented a diversity of bioprocesses related to the proteasome , glycolipid metabolism , DNA replication and repair , replication and division checkpoints , RNA splicing , microtubules , Golgi and vesicle transport , and chromatin state ( S3 Fig ) . In contrast , the 100 worst-performing terms were bioprocesses primarily related to carbohydrate , nucleotide , and coenzyme/cofactor metabolism , as well as the mitochondria , transmembrane transport , and protein synthesis and localization ( S4 Fig ) . The best-performing terms were also significantly smaller than the worst-performing ones ( 8 and 35 genes on average , respectively; rank-sum p-value < 2 . 2 × 10−16 ) , which , given the fact that we would expect the power to increase with gene set size assuming the corresponding set was still functionally coherent , suggests that our method identifies functionally specific signal . Interestingly , the relatively poor performance of many metabolism-related bioprocess terms may result from the fact that the chemical-genetic and genetic interaction screens were both performed in relatively rich medium , precluding analysis of condition-specific phenotypes for genes only required for growth in minimal medium . While the set of best-performing terms did include a diverse range of bioprocesses , the possibility of “blind spots” should always be considered when interpreting the predictions made by CG-TARGET , as they may lead to false negative results that either exclude interesting compounds ( e . g . those whose primary modes of action affect carbohydrate metabolism ) or mask potential side effects of compounds whose primary modes of action are more easily observed by this method . The prediction of perturbed protein complexes offers the opportunity to enhance the specificity of GO biological process predictions ( especially for overly-general bioprocess terms ) and investigate functional space not accessible by bioprocess annotations . As such , we investigated the potential to expand the use of CG-TARGET to the prediction of perturbed protein complexes . When CG-TARGET was applied to predict protein complex targets for the RIKEN screen data , 714 compounds were identified with at least one high-confidence ( FDR ≤ 25% ) complex prediction , 604 of which also occurred in our original set of RIKEN compounds with high-confidence bioprocess predictions . Similar , but not completely overlapping , sets of genes ( Jaccard index > 0 . 2 ) contributed to the top 5 of both bioprocess and protein complex predictions for more than one third of these compounds ( 219; 36% ) ; this suggested that the two standards possessed both shared and complementary functional information that could be used to improve predictions . We observed that protein complex predictions narrowed down less-specific bioprocess terms and enabled predictions in places where bioprocess annotations were sparser . To assess the ability to refine bioprocess prediction specificity , we mapped each protein complex to the childless bioprocess terms that completely encompassed them and looked for substantial improvements in prediction strength from the bioprocess to its protein complex “child . ” We observed several instances in which bioprocess predictions with FDR > 25% ( not high confidence ) could be converted to high-confidence predictions by refining the bioprocess term to a constituent protein complex . For example , we saw substantial gains for the following bioprocess-to-complex combinations ( sizes in parentheses ) : “mRNA polyadenylation” ( bioprocess , not high confidence; size 8 ) to “mRNA cleavage factor matrix” ( complex , high confidence; size 4 ) ; “cytoplasmic translation” ( 51 ) to “cytoplasmic ribosomal large subunit” ( 24 ) ; “vacuolar acidification” ( 14 ) to “H+-transporting ATPase , Golgi/vacuolar” ( 5 ) ; and “regulation of fungal-type cell wall organization” ( 8 ) to PKC pathway” ( 4 ) ( S2 Table ) . Importantly , 27 of the 110 compounds with high-confidence protein complex but not bioprocess predictions achieved their high-confidence status purely based on protein complex predictions that enhanced the specificity of a non-high-confidence , overlapping bioprocess prediction . Additionally , a separate set of 22 out of 110 compounds achieved high-confidence status based solely on predictions to protein complexes that did not strongly overlap with any bioprocesses ( Jaccard < 0 . 2 ) , demonstrating that the current set of protein complex annotations enabled predictions in functional space that was not well captured by a GO biological process term . Predicting perturbed protein complexes also provided the opportunity to compare our method’s performance against that of a previous , protein complex-based method called PCBA ( Protein Complex-based Bayesian factor Analysis ) [12] . PCBA was designed to infer the compound-induced activities of protein complexes ( and thus predict compound mode of action ) by linking them to observed mutant fitnesses via genetic and physical interactions . The authors highlighted six compounds in their study , five of which also possessed a high-confidence ( FDR ≤ 25% ) CG-TARGET-based protein complex prediction . For the PCBA-based mode-of-action predictions , only two of the six compounds ( benomyl and nocodazole ) could be matched to their known modes of action based on protein complex activity scores alone–the remainder required additional interpretation based on the mutants that were linked to the perturbed complexes through physical or genetic interactions ( S3 Table ) . In contrast , CG-TARGET directly generated protein complex predictions related to the known modes of action for four of the five compounds with high-confidence predictions , using only the diagnostic set of ~300 mutants ( PCBA used ~3000-mutant whole-genome profiles ) . While the two studies used different sources of chemical-genetic profiles and protein complex annotations ( which precluded more rigorous comparisons ) , these limited examples suggest that CG-TARGET performs at least comparably to PCBA and possibly better when focusing just on the protein complex scores . In addition , CG-TARGET can utilize arbitrary gene sets ( including highly-overlapping GO biological process terms ) , while factor analysis-based methods such as PCBA are generally restricted to non-overlapping gene sets due to identifiability issues [12] . Our evaluations of CG-TARGET support the premise of the method that genetic interaction profiles can be used as a tool to interpret chemical-genetic interaction profiles . However , we sought to better understand the extent to which these two types of profiles actually agree with one another , and if their systematic differences could shed light on the limits of the core assumption behind our method ( i . e . that chemicals mimic the interaction profiles of their genetic targets ) . To investigate the compatibility of chemical-genetic and genetic interaction profiles , we quantified the contribution of individual gene mutants in the chemical-genetic interaction profiles to the prediction of individual bioprocesses . For a single compound and predicted bioprocess , these “importance scores” were obtained by 1 ) computing a mean genetic interaction profile across all L2-normalized query genetic interaction profiles that possessed an inner product of 2 or higher with the chemical-genetic interaction profile and mapped to the predicted bioprocess , and 2 ) computing the Hadamard product ( elementwise multiplication ) between this mean genetic interaction profile and the compound’s chemical-genetic interaction profile . Each score could have been positive , indicating agreement in the sign of chemical-genetic and genetic interactions for a gene mutant , or negative , indicating that the interactions did not agree for that gene mutant . As such , the importance scores summarized the concordance between chemical-genetic and genetic interaction profiles , conditioned on an individual compound and a perturbed bioprocess of interest . We use the prediction of NPD4142 , a compound from the RIKEN Natural Product Depository , to the “mRNA transport” bioprocess to illustrate how the overlap between chemical-genetic and genetic interactions led to bioprocess predictions ( Fig 4A ) . A qualitative examination revealed that , indeed , NPD4142 possessed a pattern of chemical-genetic interactions similar to the genetic interactions for the query genes annotated to mRNA transport . More quantitatively and as expected , we observed that the contribution of each gene mutant to a bioprocess prediction depended on the strength of its chemical-genetic interaction with NPD4142 and the number and intensity of its genetic interactions with the mRNA transport query genes . Chemical-genetic interactions with mutants of POM152 , NUP133 , and NUP188 , which encode components of the nuclear pore that facilitate import and export of molecules such as mRNA , were the most important , followed by interactions with mutants in the Lsm1-7-Pat1 complex , which is involved in the degradation of cytoplasmic mRNA . Using this approach to assess the importance of individual mutants in the chemical-genetic profile , we globally analyzed the contribution of chemical-genetic interactions to each compound’s top bioprocess prediction ( Fig 5 ) . We performed this analysis twice: first , on all HCS compounds , and second , on a diverse subset of 130 compounds to correct for potential functional biases in the full set [14] . We present here the results from the 130-compound subset , although the results for the full set were qualitatively similar . For each compound , an average of 42% of its chemical-genetic interactions contributed to its top bioprocess prediction ( chemical-genetic interaction cutoff ± 2 . 5 , importance score cutoff +0 . 1 ) –a fraction that increased substantially ( to 78% ) when limiting the analysis to each compound’s strong interactions that contributed strongly ( chemical-genetic interaction cutoff ± 5 , importance score cutoff +0 . 5 ) . Overall , we observed that more than one-third of chemical-genetic interactions ( 1112 / 3129 ) contributed to a top bioprocess prediction ( chemical-genetic interaction cutoff ±2 . 5; importance score cutoff +0 . 1 ) . Strikingly , negative chemical-genetic interactions much more frequently contributed to a bioprocess prediction: approximately one-half ( 1071 / 2112 ) of negative chemical-genetic interactions contributed as compared to only ~4% ( 41 / 1017 ) of positive chemical-genetic interactions at the same cutoff . Furthermore , we observed differences in how the signs within chemical-genetic and mean genetic interaction profiles could disagree with each other despite the global profile similarity that led to bioprocess prediction , with positive chemical-genetic interactions contributing negatively to bioprocess predictions ( importance score cutoff < –0 . 1 ) over 10 times more frequently than negative interactions ( 1 . 9% vs . 0 . 14% ) . This trend of negative chemical-genetic interactions supporting strong bioprocess predictions was even more pronounced when restricting this analysis to strong interactions ( chemical-genetic interaction cutoff ±5; importance score cutoff +0 . 5 ) , where negative interactions comprised essentially the entire set of contributing chemical-genetic interactions ( 219 / 220 , 99 . 5% ) . These observations were also supported by analyses in which we predicted perturbed bioprocesses using only negative or positive chemical-genetic interactions , finding that negative chemical-genetic interactions were the primary drivers of bioprocess predictions and overwhelmingly responsible for their accuracy [14] . We conclude that negative interactions in chemical-genetic interaction profiles contain the large majority of the functional information necessary to predict modes of action . Negative chemical-genetic interactions also contained information reflecting general effects of chemical perturbations . Specifically , we identified nine mutant strains that exhibited strong negative chemical genetic interactions ( z-score < –5 ) yet were enriched for a lack of contribution ( importance score < 0 . 1 ) to bioprocess predictions ( hypergeometric test , Benjamini-Hochberg FDR ≤ 0 . 05; shaded region of Fig 5 ) . Manual inspection of these mutants revealed connections to the high osmolarity glycol ( HOG ) pathway , cell polarity ( cytoskeletal actin polarization , kinetochore and chromosome segregation ) , and other stress response mechanisms ( S4 Table ) . As the HOG pathway is important for the cellular response to high osmolarity and other stresses [22–24] , and repolarization of the cytoskeleton is required for cells to adapt and continue dividing after stress [25 , 26] , we hypothesize that many of these overrepresented mutants interact negatively with compounds due to an impaired ability to respond to external stress . This chemical perturbation-specific information may complement or even completely obscure the chemical-genetic signature of a compound’s primary mode of action , potentially complicating the interpretation of chemical-genetic interaction profiles using a genetic interaction network . We compared the concordance of chemical-genetic and genetic interaction profiles across multiple compounds predicted to the same bioprocess , revealing that some bioprocesses were predicted by homogenous sets of chemical-genetic interaction profiles while others were much more heterogeneous despite their predicted targeting of the same bioprocess . For example , predictions made to the “CVT pathway” ( FDR < 1% ) depended almost entirely on a suite of strong negative chemical-genetic interactions with ARL1 , ARL3 , and ERV13 , with contributions from IRS4 and COG8 ( Fig 4B ) . This uniformity in the prediction of a bioprocess is contrasted by the diversity of profiles captured within “tubulin complex assembly” predictions ( Fig 4C ) . Compounds with top predictions to this term could potentially be partitioned into three classes , divided according to strong contributions from: 1 ) CIN1/TUB3 , PAN3/CIN4 , and the SWR1 complex ( known tubulin polymerization inhibitors Benomyl and Nocodazole ) ; 2 ) CIN1/TUB3 and DSE2 ( NPD4098 and NPD2784 ) ; or 3 ) only CIN1/TUB3 ( all remaining compounds except NPD4619 ) . Interestingly , the structures of the compounds in each of the former two groups are distinct from those in the other groups , suggesting that the observed diversity in these compounds’ functional profiles is mechanistically derived from their structures .
The scaling of chemical-genetic interaction screens from tens or hundreds of compounds to tens of thousands of compounds has provided the opportunity , and the necessity , to develop better methods for interpreting the interaction profiles and prioritizing high-confidence compounds . We developed a method , CG-TARGET , to address this need and applied it in a recent study to predict perturbed biological processes for 1522 out of nearly 14 , 000 compounds screened for chemical-genetic interactions [14] . Our rigorous benchmarking of CG-TARGET showed that , in terms of accuracy , it outperformed direct enrichment on chemical-genetic interactions , and in terms of false discovery rate control , it outperformed both enrichment-based alternatives ( direct enrichment and gene-target enrichment ) by identifying at least 4-fold more compounds at FDR ≤ 25% . Multiple experimental validations have further supported the accuracy of the method and its usefulness for functionally annotating previously uncharacterized compounds , with validations of predicted tubulin polymerization and mitotic checkpoint inhibitors presented here . The companion paper describes additional experimental validations , including one performed on 67 compounds based on linking bioprocess predictions to the stage of induced arrest in an orthogonal cell cycle assay [14] . This study is , to our knowledge , the first systematic evaluation of the ability of genetic interaction profiles to interpret chemical-genetic interaction profiles at a large scale . The results of this study are encouraging , as a genome-wide compendium of genetic interaction profiles provides a much more comprehensive and unbiased resource for profile interpretation than a limited set of gold standard compounds . Aggregating the compound-gene similarities into compound-bioprocess predictions not only provided for increased statistical confidence but also allowed for direct functional annotation of compounds without direct protein targets ( e . g . DNA-damaging or membrane-disrupting agents ) . Interestingly , enrichment on compound-gene similarities performed similarly to CG-TARGET in ranking bioprocess predictions for individual compounds but performed much worse on the task of prioritizing these predictions across compounds . CG-TARGET likely excelled here because it accounts both for the chemical-genetic profile strength in compound-gene similarity calculations and for the effects of general signals that arise upon treatment with bioactive compound . These general signals could be amplified through their similarity to a large cluster of profiles in the genetic interaction network and were the specific motivation for incorporating resampled profiles into the prediction scheme . Genetic interaction-based interpretation of chemical-genetic interaction profiles has revealed broad insights into chemical function and provided interesting directions for further exploration , but some questions remain to be addressed about the limits of the technique . In the companion paper , we used the results from CG-TARGET to characterize the distribution of predicted perturbed functions for entire chemical libraries , revealing a general depletion of compound action in the nucleus and an enrichment of activity near the cell wall and membrane [14] . Additionally , we investigated the hypothesis that the profile of a compound with multiple independent modes of action would resemble a combination of distinct genetic interaction profiles , which led us to a compound whose independent predictions to cell wall and DNA perturbation were both validated ( the top 20 dual-process predictions are included as Supplementary Table 2 in [14] ) . Indeed , we observed broad compatibility between chemical-genetic and genetic interaction profiles , the overwhelming basis of which was contributed by negative chemical-genetic interactions . However , we observed exceptions to this compatibility for genes to which perturbations may reduce the ability of cells to deal with external stress . In general , the fact that chemicals may induce stresses that cannot be recapitulated with genetic perturbations represents a potential blind spot in our approach , but one that could possibly be remedied by including specific stress conditions in the compendium of profiles used for interpretation . We do note , however , that every observed chemical-genetic or genetic interaction essentially represents an increased or decreased ability to deal with a particular stress , and many of our predictions are successful because the stresses induced by genetic and chemical perturbations overlap . While we demonstrated here the ability to predict perturbed bioprocesses for compounds and prioritize the highest-confidence predictions , many further steps are required to identify lead compounds and ultimately develop molecular probes or pharmaceutical agents . Perturbing a biological process does not necessarily require perturbing a specific protein target , and as such , further refinements to our methods are needed to identify specific molecular targets ( i . e . proteins ) and prioritize the compounds most likely to perturb a small number of defined targets in the cell . We envision the use of multiple functional standards with CG-TARGET , such as biological processes and protein complexes as demonstrated here , to improve our ability to predict compound mode of action at different levels of resolution and predict the compounds that exert specific versus general effects in the cell . Different modes of chemical-genetic interaction screening can provide support in this endeavor , as heterozygous diploid mutant strains , gene overexpression strains , and/or spontaneous compound-resistant mutants can provide evidence for the direct , essential cellular target ( s ) of a compound [1 , 7] . Regardless of the limitations in predicting precise molecular targets , information about the bioprocesses perturbed by an entire library would be useful in selecting the compounds most amenable to activity optimization and off-target effect minimization in the development of a pharmaceutical agent or molecular probe . The approach described here can be translated to work in other species for which obtaining functional information on compounds would be useful . For example , genome-wide deletion collections have been developed for Escherichia coli [34] and Schizosaccharomyces pombe [35] and used to perform chemical-genetic interaction screens [36 , 37] as well as genetic interaction mapping [38–41] . Such efforts are even underway in human cell lines , enabled by genome-wide CRISPR screens [42–47] . Furthermore , future efforts to interpret chemical-genetic interaction profiles in a new species need not wait for the completion of a comprehensive , all-by-all genetic interaction network as exists in S . cerevisiae , as our work highlights the ability of a diagnostic set of gene mutants to capture functional information and predict perturbed biological processes . From the discovery of urgently-needed antibacterial or antifungal agents , to the treatment of orphan diseases or a better understanding of drug and chemical toxicity , the combination of chemical-genetic and genetic interactions in a high-throughput format , with appropriate analysis tools , offers a means to achieve these goals via the discovery of new compounds with previously uncharacterized modes of action .
Our method to predict biological processes perturbed by compounds is briefly summarized in the recent study that contains its original application to a large-scale chemical-genetic interaction dataset , generating the bioprocess predictions that are subjected to further rigorous benchmarking in this manuscript [14] . The method is more formally described here . S1 Fig and S5 Table respectively provide a schematic representation and reference for variables and symbols . At a high-level , CG-TARGET predicts the bioprocesses perturbed by compounds in three major steps ( after generating a set of randomly resampled profiles to use as a control ) . First , chemical-genetic interaction profiles are compared to genetic interaction profiles to generate compound-gene similarity scores . Second , these similarity scores are aggregated into compound-bioprocess scores , which are compared against score distributions derived from negative experimental control profiles , randomly resampled profiles , and randomization of the gene labels on the compound-gene scores . Finally , false discovery rate estimates are computed by comparing the rates , across a range of p-value thresholds , at which discoveries are made for negative control and randomly resampled profiles versus the discovery rate for compound-derived profiles .
|
Understanding how chemical compounds affect biological systems is of paramount importance as pharmaceutical companies strive to develop life-saving medicines , governments seek to regulate the safety of consumer products and agrichemicals , and basic scientists continue to study the fundamental inner workings of biological organisms . One powerful approach to characterize the effects of chemical compounds in living cells is chemical-genetic interaction screening . Using this approach , a collection of cells–each with a different defined genetic perturbation–is tested for sensitivity or resistance to the presence of a compound , resulting in a quantitative profile describing the functional effects of that compound on the cells . The work presented here describes our efforts to integrate compounds’ chemical-genetic interaction profiles with reference genetic interaction profiles containing information on gene function to predict the cellular processes perturbed by the compounds . We focused on specifically developing a method that could scale to perform these functional predictions for large collections of thousands of screened compounds and robustly control the false discovery rate . With chemical-genetic and genetic interaction screens now underway in multiple species including human cells , the method described here can be generally applied to enable the characterization of compounds’ effects across the tree of life .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetic",
"networks",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"fungi",
"model",
"organisms",
"network",
"analysis",
"experimental",
"organism",
"systems",
"tubulins",
"genetic",
"interactions",
"research",
"and",
"analysis",
"methods",
"saccharomyces",
"polymer",
"chemistry",
"computer",
"and",
"information",
"sciences",
"animal",
"studies",
"proteins",
"chemistry",
"polymerization",
"yeast",
"biochemistry",
"cytoskeletal",
"proteins",
"eukaryota",
"protein",
"complexes",
"cell",
"biology",
"genetic",
"screens",
"gene",
"identification",
"and",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"chemical",
"reactions",
"physical",
"sciences",
"saccharomyces",
"cerevisiae",
"yeast",
"and",
"fungal",
"models",
"organisms"
] |
2018
|
Predicting bioprocess targets of chemical compounds through integration of chemical-genetic and genetic interactions
|
Recent technologies like AGO CLIP sequencing and CLASH enable direct transcriptome-wide identification of AGO binding and miRNA target sites , but the most widely used miRNA target prediction algorithms do not exploit these data . Here we use discriminative learning on AGO CLIP and CLASH interactions to train a novel miRNA target prediction model . Our method combines two SVM classifiers , one to predict miRNA-mRNA duplexes and a second to learn a binding model of AGO’s local UTR sequence preferences and positional bias in 3’UTR isoforms . The duplex SVM model enables the prediction of non-canonical target sites and more accurately resolves miRNA interactions from AGO CLIP data than previous methods . The binding model is trained using a multi-task strategy to learn context-specific and common AGO sequence preferences . The duplex and common AGO binding models together outperform existing miRNA target prediction algorithms on held-out binding data . Open source code is available at https://bitbucket . org/leslielab/chimiric .
Recent high-throughput technologies like AGO CLIP sequencing [1 , 2] and CLASH ( crosslinking , ligation , and sequencing of miRNA-RNA hybrids [3] ) enable direct biochemical identification of AGO binding and miRNA target sites transcriptome-wide . The miRNA field has a strong tradition of computationally leveraging transcriptome-wide data to improve target site prediction , but the leading miRNA target prediction methods today do not exploit these new biochemical data . Here we present a systematic approach to learn both the rules of miRNA-target site pairing and a binding model of AGO’s local sequence preferences and positional bias in alternative 3’UTR isoforms in order to accurately predict miRNA-target interactions . Before it became possible to map AGO-mRNA and miRNA-mRNA interactions directly , the major advance in miRNA target prediction came from restricting to predefined classes of miRNA seed matches in 3’UTRs and training a model to predict mRNA expression changes in miRNA overexpression experiments . TargetScan was the first algorithm to introduce the strategy of correlating context features of miRNA seed sites—including flanking AU content , position in the 3’UTR , and complementarity to the 3’ end of the miRNA—with extent of target down-regulation in miRNA transfection experiments [4] . Similar observations were encapsulated in the TargetRank method [5] , and these studies established that rules of miRNA targeting could be statistically decoded from transcriptome-wide data . However , new data from AGO CLIP sequencing and CLASH challenge some of the assumptions of existing prediction strategies . These data confirm the prevalence of non-canonical target sites lacking complementarity to the miRNA 2–7 ( 6-mer ) seed region and conversely show that even exact miRNA 2–8 ( 7-mer ) seed matches are often not AGO bound [6 , 7] . Meanwhile , most target prediction methods require strong seeds to avoid false positives . For example , downloadable predictions from the most recent version of TargetScan still require either perfect 2–8 seed complementarity ( 7-mer-m8 site ) or a 2–7 seed with A across from miRNA position 1 ( 7-mer-1A site ) , although AGO CLIP data suggests that 7-mer and 8-mer seeds are found in only about half of AGO binding sites [6] . The mirSVR method [8] , which also trains on miRNA overexpression experiments , allows up to one mismatch or G:U wobble in the 6-mer seed region , but in practice few non-canonical sites are assigned even moderate scores . Therefore , current target prediction methods may focus on detecting the most effective miRNA sites at the cost of missing a large proportion of miRNA-mRNA interactions . Furthermore , training on non-physiological miRNA overexpression experiments may obscure more subtle targeting rules . A few studies have developed algorithms to resolve which highly expressed miRNAs are associated with individual AGO CLIP peaks . For example , microMUMMIE is an algorithm for analysis of AGO PAR-CLIP that uses the location of T-to-C mutations—indicative of the site of cross-linking of the RNA-binding protein to the RNA in the PAR-CLIP assay—to assign the most likely canonical seed [9] . Other methods use energy-based duplex prediction to associate miRNAs with CLIP-mapped target sequences [10–13] . In particular , MIRZA uses an unsupervised probabilistic approach to learn parameters of a duplex alignment model from AGO CLIP peaks , and the duplex model can be used to make de novo miRNA target site predictions from 3’UTR sequence [12] . Note that the MIRZA study used the term “non-canonical” to refer to sites lacking 7 or 8 nucleotides of perfect complementarity to the 5’ end of the miRNA; therefore , their reported non-canonical sites included both perfect 6-mer and many 7-mer-1A sites . ( We will use “non-canonical” exclusively for sites lacking full complementarity in the 2–7 6-mer seed region . ) More recently , MIRZA-G combined MIRZA duplex quality scores with known context features like flanking AU content and predicted secondary structure accessibility as well as conservation , once again to predict extent of down-regulation in miRNA overexpression experiments [14] . Here we present a novel model for miRNA target prediction through discriminative learning on transcriptome-wide AGO CLIP and CLASH profiles . Our goal was to learn to accurately predict biochemical miRNA-target site interactions , rather than the extent of regulation , in order to increase the sensitivity of miRNA target prediction and learn physiological targeting rules . As the CLASH protocol captures direct interactions between miRNAs and mRNAs by ligation , it provides a partially labeled training set of miRNA-mRNA interactions including many non-canonical pairings , which we combined with canonical AGO binding sites identified by CLIP . We trained one support vector machine ( SVM ) classifier to model the miRNA-mRNA duplexes and a second SVM to learn AGO’s local sequence preferences in the UTR and positional bias in 3’UTR isoforms . The duplex SVM model enables the prediction of both canonical and non-canonical pairings between miRNA and target sequences and outperforms existing methods for assignment of miRNAs to AGO binding sites . The AGO binding model is trained using a multi-task strategy to distinguish between cell type and protocol specific sequence signals and common AGO sequence preferences . The duplex SVM and common AGO binding SVM together outperform existing target prediction approaches when evaluated on held out interaction data . Our prediction method , called chimiRic , is available as open source code at https://bitbucket . org/leslielab/chimiric .
ChimiRic’s duplex model is trained on chimeric reads from CLASH data , which associates a miRNA with a target sequence via chimeric reads and can identify non-canonical binding sites , and AGO CLIP binding sites containing a 6-mer seed match ( or longer seed ) for a single highly expressed miRNA ( Fig 1A ) . In the latter case , differential AGO CLIP-seq analysis suggests that an AGO bound site that can be associated with a unique miRNA by a canonical 6-mer seed is likely a binding site for that miRNA [6] . We used CLASH [3] and AGO PAR-CLIP data [2 , 15] in HEK293 cells to train the duplex model , restricting to the top 59 expressed miRNAs in 21 miRNA seed families ( S1 Table , Materials and Methods ) . To compile the training set , sites identified by CLASH chimeric reads were required to fall within 3’UTRs , contain a sequence within an edit distance of 1 ( substitutions or indels ) from a canonical 6-mer seed match for the interacting miRNA , and also be supported by non-chimeric reads ( see Materials and Methods ) . This filtering yielded the positive training examples consisting of 1 , 727 ( miRNA , site ) pairs supported by chimeric reads , of which 1 , 228 were non-canonical interactions , together with 11 , 211 canonical ( miRNA , site ) examples from AGO CLIP sites ( Fig 1A ) . Canonical miRNA seed matches that are not AGO bound based on CLIP data , together with ( miRNA , site ) pairs where an AGO-bound site is paired with an incorrect miRNA , provided 25 , 411 negative examples . To compensate for the class imbalance , we only used a randomly sampled subset of negative examples in training ( see Materials and Methods ) . We trained a structural SVM [16] on positive and negative ( miRNA , site ) training examples to learn a model for predicting miRNA-site duplex alignments . Here , the model vector w of the SVM represents the scoring parameters for local pairwise alignment . SVM training proceeds iteratively , alternating between obtaining optimal alignments of all training examples given the current SVM parameters w and updating the model vector w given the current duplex alignments ( Materials and Methods , S1 Fig ) . The model update step involves solving the SVM large-margin optimization problem so that the discriminant scores assigned to positive and negative ( miRNA , site ) examples have the correct sign and obey margin constraints , with a hinge loss function to control margin violations ( see Materials and Methods ) . To define the local alignment scoring system and convert the alignment score into an SVM discriminant function , we used a parameterization similar to the energy-based scoring system in MIRZA , namely a match/mismatch score that depends on the position in the miRNA sequence together with the nucleotides being aligned and penalties for loop opening and for symmetric and asymmetric loop extensions ( see Materials and Methods ) . One important difference with MIRZA is that the chimiRic alignment can only start at position 1 of the miRNA if is it matched against nucleotide A , which more accurately reflects known determinants of miRNA targeting [17] . The second component of chimiRic’s scoring system is an SVM classifier that learns to discriminate the local sequence features and positional bias in 3’UTR isoforms of true AGO binding sites versus sites that contain 6-mer seed matches of highly expressed miRNAs but are not AGO-bound , as determined by CLIP data ( Fig 1B ) . Here we considered two AGO CLIP sequencing data sets , the human HEK293 PAR-CLIP data set [2] as well as a HITS-CLIP data set in activated mouse CD4+ T cells [6] . The local sequence context of the upstream and downstream 30 nt regions flanking the 6-mer seed match are represented using weighted degree kernels [18] , which encode position specific k-mers for k = 1 … 6 ( see Materials and Methods ) . The positions of 3’ ends of alternative 3’UTR isoforms were identified from a human 3’-seq tissue atlas [19] and a mouse PolyA-seq atlas [20] . For each site in human or mouse , positional information was encoded by a vector of distance values ( measured in nucleotides ) to the annotated stop codon and to the nearest mapped 3’ ends and transformed using a radial basis kernel ( see Materials and Methods ) , and the sum of the weighted degree kernels and positional radial basis kernel was used to train the SVM . In order to model differences in AGO binding preferences between the two data sets—both due to protocol differences and potentially due to cell-type specific factors influencing AGO occupancy—we used multi-task learning to train cell-type specific AGO preference models together with a common AGO binding model ( Fig 1B , Materials and Methods ) . The cell-type specific models are intended to absorb sequence signals that predict AGO binding in a context-dependent manner , while the common model can be used for target prediction in any new context . To evaluate chimiRic’s duplex model , we held out from training all HEK293 CLASH interactions for a single miRNA seed family ( positive test examples ) together with a collection of targets sites that interact with other miRNAs based on chimeric read evidence ( negative test examples ) , and we assessed whether the model could rank the held-out miRNA family’s true target sites above these other sites . For each held-out miRNA family in turn , we used chimiRic to generate and score the duplexes between miRNAs in the seed family and mRNA site sequences in the test set . We found that the duplex model could more accurately discriminate true from false interactions compared to MIRZA , an existing method for learning miRNA-mRNA interactions from CLIP data , based on area under the ROC curve ( auROC ) analysis ( Fig 2A , blue points , p < 3 . 02e-5 , signed rank test ) . Note that the original MIRZA model was trained on the same HEK293 PAR-CLIP data set as we used to train the duplex model . To further evaluate the performance on independent data sets , we then used the duplex model trained on HEK293 CLIP and CLASH data to predict miRNA-mRNA interactions supported by chimeric reads from iPAR-CLIP in C . elegans [21] and CLEAR-CLIP in mouse brain [22] . Again , chimiRic’s duplex model outperformed MIRZA for the task of ranking observed interactions for each miRNA seed family above interactions with targets sites of other miRNAs in both C . elegans ( Fig 2A , green points , p < 1 . 45e-2 , signed rank test ) and mouse brain ( Fig 2A , purple points , p < 4 . 87e-2 , signed rank test ) data sets . These results suggest that chimiRic’s miRNA-mRNA duplex model can generalize across organisms and protocols for mapping miRNA-mRNA interactions . Previous differential CLIP and CLASH studies have revealed a broad spectrum of non-canonical miRNA-mRNA interaction modes , including GU wobbles , bulges and mismatches within seed sequences , and interactions relying on 3’ base pairing instead of seed pairing [3 , 6 , 7] . In order to test whether our duplex model captures some of these known patterns of non-canonical binding , we predicted duplexes for a variety of non-canonical miRNA target sites that have been validated by luciferase assays in previous studies ( Fig 2B ) . Our model have not only correctly identified the correct interacting miRNA above the other highly expressed miRNAs , despite the lack of exact 6-mer seed matches , but also produced duplex structures representative of the previously described interaction modes , including GU wobbles , mismatches and bulges in the seed region , and complementary base pairings in the 3’ region ( Fig 2B ) . Next we combined the duplex model with the AGO binding model , which is trained to discriminate between true AGO bound sites containing 6-mer seeds for highly expressed miRNAs and sites with 6-mer seeds that are not supported by AGO CLIP read evidence , based both on local sequence context and positional bias within 3’UTR isoforms . We used a multi-task strategy to train on AGO-bound versus unbound canonical seed sites for highly expressed miRNAs in two AGO CLIP data sets , HEK293 PAR-CLIP [15] and HITS-CLIP in mouse CD4+ T cells [6] . This procedure learned both task-specific SVM models of AGO binding and a common SVM model . The task-specific SVMs may capture protocol-specific CLIP biases and/or cell-type specific AGO binding preferences . For target prediction in a new context where no CLIP data is available , the common SVM provides a “cell-type agnostic” model of AGO sequence and position preferences . To evaluate the combined chimiRic model , for each miRNA seed family , we held out all HEK293 positive target site sequences—both canonical and non-canonical sites supported by chimeric reads from CLASH as well as canonical sites with AGO CLIP read evidence that can be unambiguously assigned to the seed family—and negative site sequences , for training both the duplex and AGO binding models . We then asked how well the combined model performs at discriminating AGO-bound from unbound canonical sites relative to TargetScan [4 , 23] and mirSVR [8] , two widely used miRNA target prediction algorithms . Fig 3A shows precision-recall curves for the combined chimiRic duplex and HEK293-specific AGO binding model as well as for TargetScan and mirSVR for prediction of canonical sites for several miRNA families . Since TargetScan requires greater seed complementarity than the canonical 6-mer seed ( either 7-mer 1A or complementary at miRNA positions 2–8 ) , its overall recall of biochemically-defined sites is limited ( note that while the TargetScan 7 . 0 release discusses 6-mer seeds and non-canonical seeds [23] , only a very small fraction of sites were non-canonical in the prediction download files ) . Evaluating performance by area under the precision-recall curve ( auPR ) across held-out miRNA seed families showed that this performance advantage was significant over TargetScan ( Fig 3B , p < 1 . 91e-6 , signed rank test ) and mirSVR ( Fig 3B , p < 9 . 54e-6 , signed rank test ) . Moreover , even measuring performance up to 50% recall ( auPR50 ) , where there are still AGO-bound 7-mer sites to detect , chimiRic still outperformed TargetScan on held-out miRNAs in the HEK293 and T cell data sets ( S2 Fig ) . We then tested the combination of chimiRic’s duplex model and the common AGO binding model . Again we found that chimiRic significantly outperformed TargetScan ( Fig 3B , p < 1 . 91e-6 , signed rank test ) and mirSVR ( Fig 3B , p < 4 . 77e-5 , signed rank test ) on held-out miRNA seed families in HEK293 , with minor difference in chimiRic’s performance compared to the HEK293-specific model . Similarly , when predicting the biochemically defined target sites of held-out miRNA families in CD4+ T cells , chimiRic’s duplex model combined with either the T cell specific or the common AGO binding model outperformed TargetScan ( Fig 3C , p < 2 . 38e-7 and p < 2 . 38e-7 , signed rank tests ) and mirSVR ( Fig 3C , p < 2 . 38e-7 and p < 2 . 38e-7 , signed rank tests ) . As an independent validation , we also evaluated chimiRic’s performance in a third cellular context using two HITS-CLIP data sets in HeLa cells [1 , 7] . Again , we found that the common AGO binding model combined with duplex model had a significant advantage over TargetScan ( Fig 3D , p < 1 . 91e-5 , signed rank test ) and mirSVR ( Fig 3D , p < 3 . 29e-3 , signed rank test ) . Evaluation using auPR50 , which favors TargetScan by allowing reduced recall , still showed a significant performance advantage of the common chimiRic model over TargetScan and mirSVR in HEK293 and T cells , with a statistical tie on the HeLa cells ( S2 Fig ) . We also evaluated the performance of three additional methods , MIRZA-G [24] , MirTarget [13] and DIANA-microT-CDS [10] , all of which are trained on AGO CLIP data and provide one a single prediction score for each miRNA-gene interaction . When we compared the performance on the same HeLa data set , the common chimiRic model outperformed all three methods measured by auPR ( Fig 3E , p < 7 . 90e-4 , p < 1 . 91e-5 and p < 1 . 68e-3 , signed rank test ) , partly due to chimiRic’s better recall . When measured by auPR50 , chimiRic still achieved a statistical tie against these methods ( S2 Fig ) , showing that chimiRic’s top-ranked predictions are at least as accurate as other methods trained on AGO CLIP data sets . We also tested our performance relative to a typical evaluation of miRNA target prediction methods: predicting the extent of mRNA downregulation of miRNA targets . We evaluated the performance of TargetScan , mirSVR and chimiRic on eight miRNA transfection experiments in HCT116 cells [25] . Despite the fact that chimiRic was not trained on any expression data , the top predictions of chimiRic conferred a similar amount of regulation compared to TargetScan , while achieving better performance than mirSVR ( S3 Fig ) . Previous studies have suggested that 3’UTR miRNA target sites tend to reside near the stop codons or near the 3’ end of the transcript rather than the middle of 3’UTRs [4] . We confirmed a positional enrichment of AGO-bound sites near the stop codons ( Fig 4A , top ) and near the end of the 3’UTR compared to miRNA seeds with no AGO binding in CD4+ T cells across mouse transcripts . Additionally , for multi-UTR transcripts , we observed an enrichment of AGO-bound sites in the region upstream of internal 3’ cleavage sites ( as mapped by PolyA-seq ) that was absent for the negative site examples ( Fig 4A , top , p < 2 . 2e-16 , KS test ) . We also observed an enrichment of positive site examples ~200nt downstream of internal cleavage sites , suggesting that the resolution of the mapped 3’ ends in the mouse atlas is limited and/or that clusters of nearby 3’ cleavage sites confound the analysis . Likewise , we found HEK293 AGO binding sites enriched upstream of internal 3’ cleavage sites based on the human 3’ end atlas ( mapped by 3’-seq ) , with more modest downstream enrichment ( Fig 4A , bottom ) . These positional biases are encoded in the feature representation for the AGO binding model ( see Materials and Methods ) and lead to a significant performance improvement for the full chimiRic model ( mean auROC on held-out miRNA families of 0 . 775 without positional bias information vs . 0 . 849 in the full model , p < 2 . 38e-7 , signed rank test; S4 Fig ) . To further interpret the sequence features in the AGO binding model , we used the positional oligomer importance matrix ( POIM ) [26] approach to identify the significant positional k-mers . From the 1-mer POIMs , we observed not only high AU content flanking the miRNA seed matches in general but also specific positional signals like m1A and m8/9U ( S5 Fig ) , which are consistent with findings from previous studies [5 , 17] . Moreover , the representation allowed us to go beyond single nucleotide composition , which is the extent of sequence contextual information used in most previous miRNA target prediction methods , to explore more complex sequence features . Previous studies have suggested that various RNA binding proteins ( RBPs ) can bind to regions proximal to miRNA target sites in order to enhance or repress miRNA-mediated regulation [27–29] . Therefore , one potential explanation for the long positional k-mers that discriminate between AGO binding sequences and unbound sequences is that they correspond to the motifs of co-binding RBPs that mediate AGO occupancy . To explore this hypothesis , we matched the 6-mers from positions with top differential POIM scores to RNAcompete in vitro affinity data for a compendium of RBPs [30 , 31] . By measuring the enrichment of these k-mers in RNAcompete data across all RBPs and assessing significance relative to an empirical null model based on training SVMs on random permutations of the class labels ( see Materials and Methods ) , we found that the position-specific k-mers in upstream and downstream sequences were indeed consistent with several known RBP motifs ( Fig 4B ) . In the common AGO-binding model , we identified an AC-rich motif upstream of the seed match that matched an AGO RNAcompete experiment and has been proposed to be the miRNA-independent binding signal for Argonaute [31] . Meanwhile , in the downstream component of the common model , Pumilio was identified as the most significant RBP motif . It has been previously suggested that Pumilio has a role in regulating miRNA site accessibility of specific target genes [28 , 32 , 33] . Our analysis suggests that Pumilio may play a transcriptome-wide role in mediating AGO binding . We compared the HEK293 AGO CLIP to PUM2 PAR-CLIP in the same cell type [2] and found that 16 . 4% of AGO sites in HEK293 overlapped with PUM2 binding sites . Fig 4C shows one example of a miR-17/20/106 target site in the 3’UTR of UBNX2A together with sequence signals identified by the model . After decomposing the SVM sequence scores into positional prediction scores ( see Materials and Methods ) , we found that the positions with positive contribution overlapped exactly with the Pumilio binding motif and Pumilio CLIP coverage . In contrast , another miR-17/20/106 seed match site in the same 3’UTR was not bound by AGO and lacked significant positional k-mers from the sequence model .
We have presented an integrative model for predicting miRNA binding sites by training on sequencing assays that map biochemical interactions via AGO cross-linking and miRNA-mRNA ligation . We demonstrated that chimiRic can detect non-canonical miRNA-mRNA binding modes and significantly outperforms MIRZA for predicting the interacting miRNA for both canonical and non-canonical mRNA target sites . Moreover , chimiRic outperforms TargetScan , a leading target prediction method , for discriminating canonical seed sites that are bound by AGO from unbound sites . The feature representation of our AGO binding model exploits recent 3’-end sequencing data that identifies alternative 3’UTR isoforms and enables analysis of mRNA sequence signals in the vicinity of the miRNA binding sites , suggesting that other RBPs may collaborate with AGO to mediate miRNA-mRNA interactions . ChimiRic directly predicts miRNA targeting by learning from miRNA binding data , whereas most existing algorithms infer miRNA targets and model their efficiency using mRNA expression changes in miRNA overexpression experiments in cell culture [4 , 8] . One major issue with methods trained solely on gene expression changes is that the direct effects of miRNA regulation are confounded with secondary effects , leading to label noise in the learning problem . Since the true binding sites that mediate direct regulation are unknown in this setting , inference of miRNA targets involves “bootstrapping” from an initial set of assumptions of what constitutes a viable target . Furthermore , miRNA transfections in cell culture represent a non-physiological context for miRNA activity and may not accurately reflect endogenous targeting rules . Finally , miRNA binding can inhibit translational efficiency of target mRNAs in addition to or instead of reducing mRNA abundance [34] . While previous global studies suggest that miRNA-mediated changes at the mRNA and protein levels are correlated , these data also depend on miRNA overexpression in cell lines [35 , 36] . For all these reasons , it is possible that what we have already exhausted what can be learned indirectly from mRNA expression changes due to miRNA perturbations—and from miRNA overexpression experiments in particular—and that new AGO CLIP and CLASH technologies for mapping direct interactions are required to advance our understanding of miRNA targeting in cells . However , recent assays for mapping AGO sites and miRNA-mRNA interactions are technically difficult and present significant challenges for computational analysis and training of predictive models . CLASH and similar protocols that use RNA ligation to capture miRNA-mRNA interactions currently have very low ligation efficiency ( only ~2% of reads are chimeric ) [3 , 21] , suggesting that a large number of miRNA-mRNA interactions remain uncaptured . Some non-canonical interactions recovered by CLASH may be due to artifacts or biases in the ligation experiments , and one previous study found that incorporating chimeric reads into MIRZA did not significantly improve prediction performance [24] . Even in the more mature CLIP assays , data reproducibility is still limited and strongly affected by technical differences between various protocols ( e . g . PAR-CLIP , HITS-CLIP , iCLIP ) that produce protocol-specific biases [15] and by the potential false positives resulted from background binding [37] . In our experiments , we only trained on data sets with multiple biological replicates in order to ensure saturating coverage and to correctly label the mRNA sites as positive or negative . We further used a multi-task strategy to absorb dataset-specific differences into task-specific models and learn a common model that captures general sequence signals and positional preferences of AGO binding . Although the extent of miRNA target context-specificity remains unclear [38 , 39] , it is still possible that there are true biological differences in AGO occupancy between cell types . Indeed , even directed perturbation of a single miRNA-mRNA interaction can lead to distinct changes in functional responses in different immune cell types [40] . Ultimately , as CLIP-based technologies mature and larger data sets accrue , the algorithmic approaches we present here may reveal the RNA sequence elements and trans-acting factors that mediate cell-type specific miRNA-mRNA interactions .
Argonaute PAR-CLIP data in HEK293 cells [15] , HITS-CLIP in mouse CD4+ T cells [6] and HITS-CLIP in HeLa cells [1 , 7] were used to define the miRNA target sites in mRNA sequences . Reads from the wild-type libraries were aligned to hg19 and mm9 genome using the bwa aligner [41] . Argonaute binding sites were then identified from the coverage profile of uniquely aligned reads using a previously described peak calling algorithm [19] . Chimeric reads from CLASH in HEK293 cells [3] , iPAR-CLIP in C . elegans [21] , and CLEAR-CLIP in mouse brain [22] were also used for training ( HEK293 ) and testing ( C . elegans and mouse brain ) the duplex model . We used the list of miRNA-mRNA interactions provided in the original publications with additional filtering . Interactions were chosen according to following criteria: ( 1 ) binding sites were located in the 3’UTR; ( 2 ) binding sites contained complementary matches to the interacting miRNA 6-mer seed with edit distance of 0 or 1; ( 3 ) interactions were also supported by non-chimeric reads . 3’-seq data in human tissues [19] and PolyA-seq data in mouse tissues [20] were used to construct 3’UTR isoform atlases in human and mouse . The processing procedure was the same as previously described [19] . The array data set included gene expression changes in eight individual miRNA transfection experiments in HCT116 cells ( miR-15a , miR-16 , miR-215 , miR-17 , miR-20a , let-7c , miR-106b and miR-103a , corresponding to GEO data sets GSM156545 , GSM156546 , GSM156548 , GSM156553 , GSM156554 , GSM156557 , GSM156576 and GSM156580 ) [25] . To reduce the noise from genes with baseline expressions , we restricted our analysis to probes with signal intensities above median in the control experiments . We also only included the genes with a single potential target site of the transfected miRNA in order to simplify the analysis . The extent of downregulation was represented by the log2 fold change between 24 h post-transfection and control , while genes with multiple probes were represented by the median of all probes . For the duplex model , each example was a pair of miRNA and mRNA site sequences . If the same interaction was identified by chimeric reads in the CLASH data , then we considered the interaction to be positive . Otherwise , if the site was not interacting with the miRNA or another miRNA from the same seed family , then we considered it as a negative example . Due to the limited number of interactions identified by CLASH , we also added interactions inferred from CLIP data by assuming that Argonaute binding sites containing 6-mer seed matches interacted with the corresponding miRNAs , while sites without Argonaute binding were unlikely to interact with miRNAs . These assumptions provided another set of positive and negative examples of miRNA-mRNA interactions . For the AGO binding model , each example was a 3’UTR site matched to the 6-mer seed of one of the highly expressed miRNAs in the corresponding cell type ( HEK293: 59 miRNAs from 21 miRNA families; CD4+ T cells: 58 miRNAs from 24 miRNA families ) . If a seed match overlapped with an Argonaute binding site in the CLIP data , we identified it as a positive example for the corresponding miRNA . Otherwise , if a seed overlapped with no Argonaute CLIP reads , we considered it to be a negative example . We adapted the feature representation from MIRZA [12] to describe the duplex structures formed between interacting ( miRNA , site ) pairs . Three types of features were included in the representation: ( 1 ) the type of base pair ( GU , UG , AU , UA , GC , CG ) at each position in the alignment; ( 2 ) the bases where a loop is opened , symmetrically extended or asymmetrically extended in the duplex structure; ( 3 ) binary variables for each position in the miRNA sequence representing whether it is paired to an mRNA base or not . One major change we made to the original representation was that the only permissible base pairing of the first base in the miRNA was with an A in mRNA sequence , so that only an A across from position 1 would contribute positively to the score . This restriction is derived from the observations in previous studies [4] . We described the mRNA sites with two types of UTR features: local sequence context and global positional context . The sequence context was represented by positional k-mer features ( k = 1 , … , 6 ) from 30 nt sequences upstream and downstream of the miRNA seed match and implemented using two weighted degree string kernels [18] . Three positional context features for each site were computed as ( i ) the distance to the nearest stop codon , ( ii ) the distance to the next end of a 3’UTR isoform , and ( iii ) the distance to the previous end of a 3’UTR isoform and were renormalized with a radial basis kernel . These local sequence kernel and positional kernel were then combined by summing kernel matrices . We trained the duplex model both on ( miRNA , site ) examples directly derived from CLASH interactions and on examples with interactions inferred from CLIP based on 6-mer seed complementarity . One major advantage of the miRNA-mRNA duplex representation described above is that the model weights w can also be used as the parameters for local pairwise alignment [12]: given the feature description φ ( miRNA , site ) for a duplex alignment , the alignment score can be described by the additive scoring function w·φ ( miRNA , site ) . Therefore , by iteratively optimizing the model weights given the currents alignments and then computing the optimal alignments given current model weights , we can simultaneously optimize the duplexes and the scoring model . The initial duplex structure for each ( miRNA , site ) pair was predicted by duplexfold in the ViennaRNA package [42] , and the corresponding duplex feature vectors were then used to train a linear support vector machine ( SVM ) classifier . The model weights w were then used as local alignment parameters to update the duplex structure between the miRNA and mRNA site sequences . The same process was repeated for 12 iterations , by which point the model vector had converged , and the final duplex structures and model weights were used as the duplex model’s output . To compensate for the class imbalance , in each iteration we only used a fraction of negative examples randomly sampled from the whole set while using all positive examples . Specifically , we sampled 15 times as many CLASH negatives as CLASH positives , and the same number of CLIP negatives as CLIP positives . We applied a regular SVM classifier to the UTR kernel matrix when we trained the AGO binding model using CLIP training data from a single cell type . When we combined data sets from multiple cell types , we applied the multi-task learning approach [43] and treated the different cell types as different but related learning tasks to address the possibility of cell type specific miRNA targeting and AGO binding rules as well as protocol specific biases . We implemented the multi-task SVM as a modification to the kernel matrix: Kst ( x , z ) = ( μ+δst ) K ( x , z ) If two examples x and z belong to the same task ( in other words , two sites were from the same cell type ) , then an extra weight is added to their product in the kernel matrix to reflect the relationship . The free parameter μ controls the closeness of task-specific models to the average model , and its optimal value was determined by five-fold cross-validation . All the machine learning procedures described above were implemented with Numpy ( http://www . numpy . org ) and the Shogun machine learning tool box ( http://www . shogun-toolbox . org ) . The latest TargetScan 7 . 0 predictions for human and mouse ( context++ scores ) were downloaded from http://www . targetscan . org and mirSVR predictions for human and mouse were downloaded from http://www . microrna . org . For both methods , if any target site had multiple possible interacting miRNAs , we used the interaction with the highest prediction score . Predictions for human genes from MIRZA-G ( seed-MIRZA-G-C variant ) , DIANA-microT-CDS and MirTarget were downloaded from http://www . clipz . unibas . ch/index . php ? r=tools/sub/mirza_g , http://diana . imis . athena-innovation . gr/DianaTools/index . php ? r=microT_CDS/index and http://mirdb . org . Since these methods provide one single prediction score for each miRNA-gene interaction , for this comparison we also simplified our predictions by using the highest score for genes with multiple sites for the same miRNA . In order to interpret the local sequence context features captured by the learned AGO binding SVM , we computed the positional oligomer importance matrices ( POIMs ) [26] for the upstream and downstream weighted degree string kernels , which represent the positional k-mer features enriched in Argonaute binding sequences . For both POIMs , we chose the positional 5-mer or 6-mer with the highest differential POIM weight and used 15 k-mers with highest POIM weights from that position to represent the most significant motifs within the positive sequences . We then matched them to potential RNA binding protein motifs identified by RNAcompete assays [30 , 31] . Normalized array probe intensities for 208 RBPs were downloaded from the supplemental websites ( http://cisbp-rna . ccbr . utoronto . ca; http://www . cs . toronto . edu/~taehyung/gr_ago . html ) . For each RNAcompete experiment , we selected the top 1000 probe sequences with highest intensity and performed a one-sided Wilcoxon rank-sum test comparing the probes with and without the top k-mers to test the significance of enrichment . Due to the biased nucleotide content near miRNA targets , it is necessary to estimate false discovery rates ( FDRs ) for the statistical tests . For each RNAcompete experiment , we generated an empirical null distribution of p-values by training the AGO-binding SVM models 1 , 000 times with randomly permutated labels , extracting the top POIM k-mers with the same k and position as in the real model , and testing the enrichment of the top k-mers from the random models within the top probes . The FDRs were then computed by converting the enrichment p-values from the real model to empirical p-values from the 1 , 000 rounds of permutations . To better relate enriched k-mer signals to the biological context , we also filtered out RBPs with no homologs in mouse and human according to cisBP-RNA ( http://cisbp-rna . ccbr . utoronto . ca ) or with low mRNA abundance according to RNA-seq data in the same cell types [3 , 6] . Of the remaining RBPs , we considered the top 5 as ranked by FDR as the ones with potential motif enrichment near the miRNA target sites . To examine the contribution of positional k-mer features at specific binding sites , we decomposed the SVM score by summing up the SVM weights for all k-mers from the same position in the upstream/downstream sequence model . One example was visualized in Fig 4C to show the overlap between RBP binding and the corresponding sequence signals .
|
MicroRNAs ( or miRNAs ) are a family of small RNA molecules that guide Argonaute ( AGO ) to specific target sites within mRNAs and regulate numerous biological processes in normal cells and in disease . Despite years of research , the principles of miRNA targeting are incompletely understood , and computational miRNA target prediction methods still achieve only modest performance . Most previous target prediction work has been based on indirect measurements of miRNA regulation , such as mRNA expression changes upon miRNA perturbation , without mapping actual binding sites , which limits accuracy and precludes discovery of more subtle miRNA targeting rules . The recent introduction of CLIP ( UV crosslinking followed by immunoprecipitation ) sequencing technologies enables direct identification of interactions between miRNAs and mRNAs . However , the data generated from these assays has not been fully exploited in target prediction . Here , we present a model to predict miRNA-mRNA interactions solely based on their sequences , using new technologies to map AGO and miRNA binding interactions with machine learning techniques . Our algorithm produces more accurate predictions than state-of-the-art methods based on indirect measurements . Moreover , interpretation of the learned model reveals novel features of miRNA-mRNA interactions , including potential cooperativity with specific RNA-binding proteins .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"sequencing",
"techniques",
"cell",
"physiology",
"cell",
"binding",
"rna-binding",
"proteins",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"gene",
"regulation",
"immunology",
"neuroscience",
"micrornas",
"mathematics",
"forecasting",
"statistics",
"(mathematics)",
"artificial",
"intelligence",
"sequence",
"motif",
"analysis",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"computer",
"and",
"information",
"sciences",
"white",
"blood",
"cells",
"sequence",
"alignment",
"animal",
"cells",
"proteins",
"mathematical",
"and",
"statistical",
"techniques",
"gene",
"expression",
"support",
"vector",
"machines",
"t",
"cells",
"molecular",
"biology",
"biochemistry",
"rna",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"non-coding",
"rna",
"cognitive",
"science",
"statistical",
"methods",
"machine",
"learning"
] |
2016
|
Learning to Predict miRNA-mRNA Interactions from AGO CLIP Sequencing and CLASH Data
|
During the last few decades , several studies have analyzed and described knowledge , attitudes , and practices ( KAP ) of populations regarding dengue . However , few studies have applied geometric data analytic techniques to generate indices from KAP domains . Results of such analyses have not been used to determine the potential effects of sociodemographic variables on the levels of KAP . The objective was to determine the sociodemographic factors related to different levels of KAP regarding dengue in two hyper-endemic cities of Colombia , using a multiple correspondence analysis ( MCA ) technique . In the context of a cluster randomized trial , 3 , 998 households were surveyed in Arauca and Armenia between 2012 and 2013 . To generate KAP indexes , we performed a MCA followed by a hierarchical cluster analysis to classify each score in different groups . A quantile regression for each of the score groups was conducted . KAP indexes explained 56 . 1% , 79 . 7% , and 83 . 2% of the variance , with means of 4 . 2 , 1 . 4 , and 3 . 2 and values that ranged from 1 to 7 , 7 and 11 , respectively . The highest values of the index denoted higher levels of knowledge and practices . The attitudes index did not show the same relationship and was excluded from the analysis . In the quantile regression , age ( 0 . 06; IC: 0 . 03 , 0 . 09 ) , years of education ( 0 . 14; IC: 0 . 06 , 0 . 22 ) , and history of dengue in the family ( 0 . 21; IC: 0 . 12 , 0 . 31 ) were positively related to lower levels of knowledge regarding dengue . The effect of such factors gradually decreased or disappeared when knowledge was higher . The practices indexes did not evidence a correlation with sociodemographic variables . These results suggest that the transformation of categorical variables into a single index by the use of MCA is possible when analyzing knowledge and practices regarding dengue from KAP questionnaires . Additionally , the magnitude of the effect of socioeconomic variables on the knowledge scores varies according to the levels of knowledge , suggesting that other factors might be influencing higher levels of knowledge .
Dengue is an endemic or epidemic disease in countries located in the tropics [1] , and approximately 40% of the world’s population is at risk of suffering dengue [2] . In Latin America , the number of cases has increased over the last three decades; it has grown from a low dengue-endemic to a hyper-endemic state in the majority of the region [3] . In 2010 , Colombia faced the most important epidemic in the history of the country and reported the second-highest incidence in the Americas ( 157 , 203 cases ) . Since then , the country has reported the third and second-highest incidences on the continent [4 , 5] . Given the absence of a licensed and widely distributed tetravalent vaccine against dengue , integrated vector management ( IVM ) has played an important role in the control of the disease [3] . IVM is a process to optimize resources for vector control , and it has identified the assessment of local KAP as crucial in designing preventive interventions adapted to any context [6] . Since 2000 , we found approximately 51 descriptive and analytical KAP studies regarding dengue in several countries , most of these papers are descriptive . The questionnaires used in such KAP studies have ranged between 13 [7] and 75 questions [8] and often result in a large number of dichotomous variables that are difficult to synthesize . Thirteen out of 48 revised studies have generated a single KAP variable through the construction of indexes based on what researchers consider a correct answer of each Knowledge , Attitude and Practice domain [8–19] . And only one has implemented a principal correspondence analysis ( PCA ) to summarize this information [20] . However , none of these fourteen studies with indexes has analyzed KAP data to identify the patterns of responses in each of its domains . And more importantly , the results from these analyses have not been used to determine the potential effects of sociodemographic variables on the levels of such indices of KAP ( KAP index as a dependent variable ) . KAP surveys are widely used in the broad context of public health , not only for research [21] but also for planning and intervention design [22 , 23] . Additionally , they are used to assess potential participation in prevention strategies of multiple diseases [22 , 23]; to evaluate the effectiveness of several public health interventions ranging from cardiovascular diseases [24 , 25] to tropical neglected diseases [26] and also to assess healthcare services in the hospital environment [27 , 28] . MCA is a descriptive method that allows the analysis of multiple categorical variables . It is widely used to generate assets and wealth indexes [29–32] . Recently , it has been used in the analysis of behavioral variables in HIV [32 , 33] , healthy lifestyles [34 , 35] , and hantavirus [36] , amongst others . The generation of indices or scales from self-reported information has been an increasing need in social and public health sciences and new methods and theories have become increasingly popular during the last years , this is the case of latent variable analysis , factor analysis and item response theory amongst others [37] In this context , KAP surveys provide a large amount of categorical data , and MCA allows the linking of separate sets of data efficiently for finding comparable trends between them [38] . The objective of this study was to determine the sociodemographic factors associated with certain KAP levels regarding dengue in two hyperendemic cities in Colombia using the MCA technique .
KAP surveys were collected between December 2012 and April of 2013 in the cities of Colombia , Arauca and Armenia . Arauca , the capital city of the Arauca Department , has 85 , 994 inhabitants and is located on the border with Venezuela at 125 meters above mean sea level ( MAMSL ) . It has a mean temperature of 30°C . Armenia , with 293 , 614 inhabitants , is the capital city of the Quindío Department and is located in the center of the country at 1 , 483 MAMSL , with temperatures ranging from 18°C to 29°C . We surveyed 3 , 998 households in the context of a cluster-randomized trial , following the same methods of Quintero et al [39] where a grid was overlapped in a satellite image of the two cities . Areas with empty land and non-residential zones were excluded . Of the remainder , 20 squares were randomly selected in each city , and 100 households were surveyed in each square beginning by the south-west corner of each square; the group of houses was called cluster . Personnel from the health authorities of both cities visited each household and invited the responsible adult available to participate in the study . In the case of absence two additional visits in different schedules were done . If contact was not possible after three visits the household was replaced by the contiguous household ( Response rate: 99 . 95% ) . The KAP questionnaire was based on a review of published studies using KAP surveys between 2001 and 2012 and on a review of the questionnaires provided by the authors of such studies . We developed our KAP survey using a combination of questions from various KAP questionnaires . The new questionnaire was then piloted in a village near one of the study sites . After adaptation to the language in the local community , the 84-question survey was applied to each household using the mobile application e-mocha® , created by the Center for Clinical Global Health Education at the Johns Hopkins School of Medicine . The KAP survey had five sections: sociodemographic and gender decision-making information and knowledge , attitudes , and practices data ( S1 File ) . The sociodemographic section collected information about age , sex , education , income , number of persons per household , dwelling materials ( floors and walls ) , migration , and access to public services . Given the already documented difficulties to capture household wealth with self-reported income [18] , we used an additional measurement of socioeconomic strata that is used in Colombia to classify areas in the cities on a scale from 1 ( lowest ) to 6 ( highest ) and it is usually utilized to grant subsidies to the lowest-income population ( strata 1 ) and to charge differential fees for public sanitation services [40 , 41] . The gender decision-making segment inquired about who decides about the health care of their own and others , daily expenses , large expenses , and household maintenance . The gender decision-making questions were extracted from the women’s module of the Demographic Health Survey [42] . Knowledge was defined as the understanding of a specific phenomenon , in this case , the means of transmission , symptoms , and means of prevention of dengue . An attitude refers to the organization of beliefs around a concept that predisposes to act in some specific manner . In this study , we asked for the severity of dengue and the repercussion for a case in the community , among others ( e . g . Does a case of dengue in this community affects this household ? ” ) . Practices relate to a group of actions to ameliorate or trigger a specific outcome in health , in this case , actions toward the prevention of vector breeding sites [43 , 44] . All participants in this study provided oral and written informed consent before conducting the survey . The Fundación Santa Fe de Bogotá’s ethics committee , in compliance with all Colombian regulation governing the protection of human subjects , approved the protocol and the instruments of this study as recorded in the minutes of the meeting held on November 19 , 2012 . MCA was conducted to summarize the information of the categorical variables of the KAP survey into three scores ( knowledge , attitudes , and practices ) using the methods described by Kohn and Le Roux and Rouanet [38 , 45] . Unlike PCA , variables in this analysis do not need to follow a normal distribution , which makes MCA an appropriate approach for KAP variables , since most of them are categorical . We described all the dimensions extracted from the MCA; however , we chose the dimension with higher inertia for our analysis . The first step of this process was the generation of a weight-per-answer option for each domain ( knowledge , attitudes , and practices ) . To assign a score to each person according to their particular set of answers , a linear combination of the weights was done . For ease of interpretation , the scores were rescaled to be greater than or equal to 1 , where 1 is the minimum value of the index . These three scores were classified into different groups using hierarchical cluster analysis following the agglomerative method by average linkage presented by Kaufman and Rousseeuw [46 , 47] . Afterwards , we determined the number of groups through the Duda , Hart , and Stork stopping rules index [48] . Finally , a characterization of the most frequent answers in each group was done to determine the KAP profiles ( low , medium , and high ) . In some cases , this process was not possible , given the heterogeneity of the answers . A regression analysis was conducted to assess possible sociodemographic and gender determinants of the KAP scores . We considered fixed effects to account for the correlation within each cluster generated by unobservable variables . In the case of a non-normal distribution of the index , we did two-quantile regressions , considering fixed effects by clusters . We used STATA 13 for data depuration and analysis [49] .
We used MCA to calculate a score for each KAP domain; the first dimension explained 56 . 13% , 79 . 66% , and 83 . 16% of the variances , respectively . The knowledge score had inertia of 0 . 01 ( 66 variables ) , the average score was 4 . 24 ( std . dev . = 1 ) , and the maximum value was 6 . 96 . The attitude score had inertia of 0 . 122 ( 17 variables ) , with a mean score of 1 . 40 ( std . dev . = 1 ) , and the maximum value was 7 . 02 . The practices score had inertia of 0 . 05 , an average of 3 . 18 ( std . dev . = 1 . 1 ) , and a maximum value of 10 . 68 ( Fig 1 ) . As a result of the hierarchical cluster analysis , we determined five profiles in the knowledge domain according to the score generated using MCA . Profile 1 was characterized by participants not having heard about the disease and no reported knowledge about any feature of the means of transmission , clinical presentation , characteristics of Aedes aegypti , or prevention measures . Profile 2 entailed individuals who despite having heard about dengue and its means of transmission did not know about preventive measures or any other aspect of dengue or the vector . Profiles 3 and 4 included individuals who had knowledge about oviposition places ( any stagnant water ) and means of transmission . Additionally , individuals assigned to profile 4 named more constitutional symptoms , while those in profile 3 named more hemorrhagic symptoms ( such as petechiae , epistaxis , etc . ) . Profile 5 was characterized by a high knowledge about the means of transmission and recognition of the white-striped legs of the vector ( Table 2A ) . Attitude analysis generated nine profiles that did not show specific patterns per profile in the components of attitudes but could be grouped into two types: the individuals who thought that dengue is important to the community and to them , and the ones who did not . The remaining variables such as considering dengue as a serious disease and that dengue is an issue for the community and for them were evenly distributed across profiles . However , the first group accounted for 95% of the individuals , revealing that there was not enough variance between the groups . Moreover , no meaningful pattern was identified when categorizing into quartiles . For this reason , this domain was excluded from the subsequent phases of the analysis . Practices scores resulted in seven profiles . Profiles 1 and 2 were characterized by poor prevention practices against vectors , such as no coverage of water containers or water treatment , no education to other members of the household , and a low frequency of emptying water from containers more than seven days , regardless of its capacity . Persons who did not cover or add chemical substances to water containers , but who emptied water containers , were part of profile 3 , and the best practices corresponded to profiles 4 , 5 , 6 , and 7 ( Table 2 ) . The distribution of the profiles followed a descendant order , whereby the smallest score was in profile 1 and the highest in profile 7; for this reason , practices scores were treated as ordinal variables .
To our knowledge , this is the first time that MCA has been applied to analyze knowledge , attitudes , and practices regarding dengue to generate an ordinal score from a data set with several categorical variables . The impact of years of education and history of dengue among households increased dengue knowledge only among low- and medium-level knowledge profiles . The effect of more than one person reporting housekeeping in the same household as their principal occupation had a negative effect on the middle- and high-level knowledge scores . Furthermore , decision making about family health care shared by men and women increased the score of knowledge at any level . Finally , practices scores were not related to any of the measured sociodemographic or gender decision-making variables . Age and education have also been identified as the only sociodemographic variables associated with more knowledge about dengue by other studies in Thailand [50 , 51] , Malaysia [52] , Cuba ( only age ) [20] , Indonesia [13] , and Jamaica ( only education ) [53] . Two studies , one in Laos and one in Malaysia , reported no statistical association between age or education with the ability to name more than one symptom of dengue [17 , 54] . In our study , we found that education levels have a positive relationship with the improvement of knowledge up to a certain point , and it does not show an effect on the higher score level . The described relationship might be caused by the decrease in variance of the households with the highest knowledge score , which does not allow us to detect differences between education levels . However , this finding can also be another manifestation of the previously documented “base education hypothesis , ” in which the effect of education is not linear—beyond 12 years of formal education attained ( corresponding to a high school level ) , it does not seem to affect other outcomes in health [43] . In this case , higher levels of outcome cannot be achieved solely by increasing education , and it is suggestive that the underlying mechanisms for having detailed knowledge of dengue , such as the color of the mosquito’s legs , are different from those for middle and low levels of knowledge . Examples of such mechanisms , which are in accord with other results of this study , are the levels of empowerment of the family and their access to information , as hypothesized by Cutler and Oreopolus in their study about mortality [55 , 56] . Studies in other settings have found an association between socioeconomic status ( SES ) and knowledge; however , there are several ways to measure SES and knowledge . While Castro et al . used a household asset score [20] in Cuba , and Itrat used monthly income [14] in Pakistan , in this study , we used a socioeconomic stratification system utilized by the government that entails characteristics of the neighborhood and income among others [40] . The lack of a significant association in this study could be due to accounting for the confounding effect of clustering in the relationship between SES and the degree of knowledge regarding dengue . SES , as measured in the study , is clustered in neighborhoods , and other studies in the area have shown that knowledge of dengue is also clustered by neighborhoods [57] . In our study , having more than one individual within a household reporting “housekeeping” as his or her main activity during the 10 days prior to the survey indicated a negative effect on the knowledge score . This suggests that there are higher levels of informal occupations that could not be captured in the questionnaire [58] . Informality is often associated with households facing poorer economic conditions , which creates difficulty in collecting accurate data . The observed relationship with the knowledge score is identifying a different component of SES that is not captured by income or education . Further exploration of the conditions of the population that reported more than one housekeeper should be explored . The findings of this study highlight the role of joint decision making between men and women in the family’s health care as a factor that contributes to the knowledge of dengue and its transmission . Past studies have suggested the need to approach the role of gender in the distribution of household chores and its relationship with dengue [57 , 59] . For this study , gender roles were approached from a micro sociological perspective , in which the dynamics on a small scale ( families and couples ) can be observed through decision-making processes [60] . Our results evidenced that shared decision-making processes between men and women play a significant role in the acquisition of knowledge about dengue; this finding has also been observed in contraception [61–63] and malaria [64] studies , in which such an effect is explained by a higher capacity of communication and negotiation . It seems that these two elements serve as a mechanism for consolidation , providing a better understanding of dengue . These findings and its congruence with other health outcomes pave the road to further exploration of the mechanisms in which joint decision making improves knowledge and empowerment within the household across the health spectrum [63] . Although many studies have conducted KAP surveys , few studies have addressed the question of the associated factors to preventive practices for dengue , and most of them have described these practices or their association with immature forms of the vector [54 , 65 , 66] . Despite this , our findings of a lack of significant association between practices score and sociodemographic characteristics are also found in other studies ( in Jamaica , Cuba , and Vietnam ) in which the authors suggest that cultural factors could lead to certain practices [20 , 53 , 67 , 68] . This hypothesis has been addressed by other studies and is a growing interdisciplinary field [69] . The main recommendation of the World Health Organization ( WHO ) and the Pan American Health Organization ( PAHO ) is to control the immature forms of Aedes aegypti for the recent Zika and Chikungunya outbreaks in the Americas [70 , 71] . This study suggests the need for further assessment of the determinants of the practices of vector control that move beyond sociodemographic factors . Moreover , it provides an additional tool for tackling the routine questionnaires performed during vector control campaigns . Since this is a cross-sectional analysis , one limitation is the impossibility of establishing the temporality of the relationships described . Moreover , even when considering fixed effects that allow controlling for the correlation between households of the same cluster , it is not possible to control for unmeasured confounding variables that vary over time , such as seasonal preventive interventions in some neighborhoods or unequal access to media that could confound the effect of sociodemographic characteristics on knowledge scores . Even though selection bias is a possibility , we think there are mainly three reasons for not thinking this will affect our results . The first reason is mainly because we think that since we were assessing household behavior there was no better informant than the housewife itself , most of the times when asking other person in the house they would refer or even ask the housewife about some of the practices . The second reason is that 30% of the women reported working rather than housewifery . This indicates that recruitment time also allowed us to have information about women whose main activity was different from housewifery . Finally , when exploring other studies in which KAP about dengue was done 5 out 8 reported more or equal to 50% of its participants as housewives [18 , 72–75] and only 3 reported a proportion of less than 20% [51 , 52 , 54] . This makes us think that this might be a characteristic of the type of survey that we are doing rather than a bias . In spite of the previous literature search and collection of most of the KAP questionnaires applied in the region for the development of the survey , comparability with other studies was a challenge . It would be helpful to generate a standardized and validated KAP questionnaire that allows comparisons between countries and across time . The use of factor analysis would facilitate the validation of such tools [37] , however it is of crucial importance to reach a consensus regarding the definition of each of the domains and what each knowledge , attitude and practices construct means . Additionally , it is necessary to establish a more suitable way to address attitudes , given that the heterogeneity of the responses in the context of a survey does not allow generalizations in the study population , our results provide evidence that KAP surveys have important measurement limitations . As discussed previously by Launiala [21] the measurement of attitudes via surveys is a sensitive topic , independent of the health issue of interest [76] . Measurement constrains such as respondent bias e . g . faulty recall and social desirability [77] question the validity of this KAP surveys and raise concerns about the possibility of measuring attitudes through surveys . Qualitative approaches such as interviews and direct observation may be more adequate as they allow rapport and buffer cultural barriers between the researchers and the respondents . Our results show that in the context of our study the attitudes domain cannot be summarized into one or two variables because of the heterogeneity and discordance in the data collected . Further research in this issue may provide evidence of these patterns in the context of other public health issues , which will vary given for example socio-cultural norms about the health outcome of interest . A mixed methods approach may be ideal as a methodological strategy to triangulate information about a culturally and socially sensitive topic [78] . MCA , besides from being a generalization of correspondence analysis ( CA ) , can be an adequate data procedure to reduce and summarize a large number of categorical variables into one ordinal variable , where the weighting process is due to a maximization of the overall correlation structure . It is helpful to understand the factors that might contribute to different levels of KAP in the community beyond the traditional descriptive analysis . Additionally , it can also be used as a tool to identify ways to improve the questionnaire and the classification of individuals into categories . Finally , given the broad use of KAP surveys in many aspects of public health such as research , planning and evaluating interventions across different issues in health , MCA becomes a useful tool to analyze the vast amount of data collected with a KAP questionnaire by the creation of one index , optimizing interpretation and usefulness . This study allowed us to identify multiple research opportunities , including the further use of this method to determine what levels of knowledge are associated with pupal indexes , to validate KAP surveys in different populations , to conduct further reliability studies , and to implement an abbreviated version of the method . Since the current evidence about the drivers of preventive practices against dengue is not conclusive , further exploration of such factors would help policy makers to understand and thus promote them in the population at risk . In conclusion , MCA is a useful tool for the analysis of KAP surveys . In regard to dengue , age and education are the only socio demographic factors associated with lower or mid levels of knowledge , whereas collective decision-making processes in the household are positively related to high levels of knowledge . No sociodemographic factors were associated with practices .
|
Local vector control programs , as part of their routine activities , often use Knowledge Attitudes and Practices ( KAP ) surveys to guide their dengue control strategies . Usually , these questionnaires are extensive and result in a large amount of data that is difficult to analyze and summarize . This study uses an analytical approach to summarize the results of these types of questionnaires about dengue and subsequently assesses the effect of sociodemographic factors . Our results suggest that Multiple Correspondence Analysis is a useful statistical technique to summarize KAP survey information . Age and higher levels of education are related to more reported knowledge about dengue , but these effects are seen only among people in groups with low and middle knowledge of dengue , according to their KAP knowledge scores . Finally , when decisions about the family healthcare are made jointly by male and female members of the household , knowledge about the disease and means of transmission improves at all levels of knowledge ( low , medium , high ) . Preventive practices regarding dengue , specifically against adult and immature forms of the mosquito that transmit the disease do not seem to be related to sociodemographic factors . This study provides an alternative way to more effectively analyze the results of KAP in a routine setting .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"classical",
"mechanics",
"decision",
"making",
"neuroscience",
"health",
"care",
"preventive",
"medicine",
"research",
"design",
"cognition",
"surveys",
"infectious",
"disease",
"control",
"questionnaires",
"research",
"and",
"analysis",
"methods",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"physics",
"socioeconomic",
"aspects",
"of",
"health",
"inertia",
"survey",
"research",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"cognitive",
"science",
"motion"
] |
2016
|
KAP Surveys and Dengue Control in Colombia: Disentangling the Effect of Sociodemographic Factors Using Multiple Correspondence Analysis
|
The possibility of HIV-1 eradication has been limited by the existence of latently infected cellular reservoirs . Studies to examine control of HIV latency and potential reactivation have been hindered by the small numbers of latently infected cells found in vivo . Major conceptual leaps have been facilitated by the use of latently infected T cell lines and primary cells . However , notable differences exist among cell model systems . Furthermore , screening efforts in specific cell models have identified drug candidates for “anti-latency” therapy , which often fail to reactivate HIV uniformly across different models . Therefore , the activity of a given drug candidate , demonstrated in a particular cellular model , cannot reliably predict its activity in other cell model systems or in infected patient cells , tested ex vivo . This situation represents a critical knowledge gap that adversely affects our ability to identify promising treatment compounds and hinders the advancement of drug testing into relevant animal models and clinical trials . To begin to understand the biological characteristics that are inherent to each HIV-1 latency model , we compared the response properties of five primary T cell models , four J-Lat cell models and those obtained with a viral outgrowth assay using patient-derived infected cells . A panel of thirteen stimuli that are known to reactivate HIV by defined mechanisms of action was selected and tested in parallel in all models . Our results indicate that no single in vitro cell model alone is able to capture accurately the ex vivo response characteristics of latently infected T cells from patients . Most cell models demonstrated that sensitivity to HIV reactivation was skewed toward or against specific drug classes . Protein kinase C agonists and PHA reactivated latent HIV uniformly across models , although drugs in most other classes did not .
The possibility to achieve HIV eradication has been limited , at least in part , by the existence of latently infected cellular reservoirs [1]–[3] . The major known cellular reservoir is established in quiescent memory CD4+ T cells , providing an extremely long-lived set of cells in which the virus can remain transcriptionally silent [1]–[3] . Reactivation of latent viruses followed by killing of the infected cells has been proposed as a possible strategy ( “shock and kill” ) to purge the latent reservoir [4] . Studies to examine the control of HIV latency and potential reactivation have been hindered , however , by the small numbers of latently infected cells in vivo and the absence of known phenotypic markers that can distinguish them from uninfected cells . In this setting , cell-line models of latency have been very useful due to their genetic and experimental tractability . Major conceptual leaps have been facilitated by the use of latently infected T cell lines [5]–[10] , including the ability to conduct genetic screens [11] . On the other hand , latently infected cell lines are limited by their cycling nature and inherent mutation in growth controls , and the clonal nature of the virus integration sites . Such transformed cell lines lack the ability to differentiate and naturally oscillate between phases of quiescence and active proliferation in response to biological signals . Because of these limitations , a number of laboratories have recently developed primary cellular models of HIV-1 latency that capitalize on specific aspects of the T cell reservoir , found in vivo ( reviewed in references [12]–[14] ) . These newer models afford investigators the ability to easily and rapidly study proposed mechanisms governing latency and to evaluate novel small molecule compounds for induction of viral reactivation . One significant complication , associated with the present variety of available latency models , is that notable differences exist among the cell model systems . Disparities relate to: the T-cell subsets being represented; the cellular signaling pathways that are capable of driving viral reactivation; and the genetic composition of the viruses employed , ranging from wild-type to functional deletion of multiple genes . Additional differences reside in the experimental approaches taken to establish latent infection in these primary cell models , which involve either infection of activated cycling cells that are later allowed to return to a resting state [15]–[19] , or direct infection of quiescent cells [20] , [21] . Because of such system variables , screening efforts in specific cell models with identified drug candidates for “anti-latency” therapy often fail to reactivate HIV uniformly across the different models . Therefore , the activity of a given drug candidate , demonstrated in a particular cellular model , cannot predict reliably the activity that will be seen in other cell model systems or in infected patient cells , tested ex vivo . The current situation in this research field represents a critical knowledge gap that is adversely affecting our ability to identify promising treatment compounds and their associated molecular mechanisms and is hindering the advancement of drug testing into relevant animal models and ultimately , human clinical trials . The present work represents a broad collaborative effort to compare and contrast induction of HIV reactivation across a battery of well-characterized cell models of viral latency , employing a highly coordinated and standardized testing approach . This work is based on the premise that it is unlikely that a single in vitro cell model can completely recapitulate the biological properties of the latent reservoir in vivo , let alone reflect accurately the response characteristics of infected patient cells ex vivo . Therefore , it is important to define both the common and unique properties among the available cell models of HIV latency in order to design a rational approach to employ such models in the identification of valid candidate drugs to induce HIV reactivation . Examples of how such an approach also can inform the underlying mechanistic actions of experimental compounds are available in the field . For instance , in the latency model developed by Bosque et al . [15] , the derived central memory CD4+ T cells ( TCM ) are highly responsive to stimuli that activate the nuclear factor of activated T-cells ( NFAT ) ; on the other hand , virus reactivation from J-Lat clones [8] tends to be highly responsive to stimuli that activate the nuclear factor kappa of B cells ( NFκB ) , such as protein kinase C ( PKC ) activators and tumor necrosis factor-alpha ( TNF-α ) . Although the use of these two model systems would predictably yield different types of hits during a compound library screen , it is important to note that known compounds , which signal through either of these activation pathways , are capable of reactivating HIV replication in latently infected CD4+ T cells from patients ex vivo , and by inference , perhaps in vivo . To begin to understand the biological characteristics that are inherent to each model of HIV-1 latency , we compared the properties of six models ( Table 1 ) , to those obtained with a standard viral outgrowth assay using patient-derived infected cells [1] , [22] . As no specific denominations have been assigned to these models , we have for simplicity referred to them by the name of the senior investigator in whose laboratory they were developed . They included the following ( details are provided within the Methods section ) : The Greene laboratory model [23] is a modification of the original O'Doherty model of latency [20] and establishes HIV infection directly in quiescent primary CD4+ cells , using spinoculation delivery of virus . Replication-competent NL4-3 reporter virus is used , which contains Luciferase in the nef reading frame ( Δnef/luciferase ) . After a short 3 day-culture , induction of provirus activation from latency is performed in the presence of integrase inhibitor to prevent viral spread and the contribution of any unintegrated viral species . Quantification of HIV replication by Luciferase expression is population-based . While only approximately 5–10% of the culture contains latently infected cells , this assay permits the generation and analysis of test compounds within 6 days . The model developed by Lewin and colleagues uses exposure of primary resting CD4+ T cells to chemokines that bind to receptors CCR7 , CXCR3 or CCR6 to effectively establish infection with wild-type NL4-3 virus [21] , [24] . Incubation with the chemokines does not cause significant cellular activation , but induces changes in the cellular actin cytoskeleton , which allows for efficient virus nuclear localization , integration , and establishment of latent infection [24] . Treatments to reactivate virus are followed by co-culture with amplifying feeder cells . Productive HIV replication is determined on a total population basis by quantification of soluble reverse transcriptase ( RT ) activity released into culture . The Planelles model [14] , [15] establishes viral latency in cultured primary CD4+ T cells that have been differentiated by TCR stimulation in the presence of TGF-β , and αIL-4 and αIL-12 monoclonal antibodies into a non-polarized subset , representative of central memory cells ( TCM ) [14] , [25] . Spinoculation with a packaged env defective NL4-3 clone establishes a single round of infection in the majority of the cells that transition into latency . Induced reactivation of HIV is monitored on a per-cell basis , using staining and flow cytometry detection for intracellular Gag ( p24 ) expression . The Siliciano model [17] uses a two-step derivation of latency in cultured primary CD4+T cells , isolated from peripheral blood . In the first step , cells are TCR stimulated , transduced with the EB-FLV lentiviral vector , for constitutive expression of Bcl-2 , expanded in culture with IL-2 and allowed to return to a resting state . In the second step , the cells are reactivated and infected with a trans-packaged , replication defective NL4-3 GFP-reporter virus clone ( NL4-3-Δ6-drEGFP ) . After 3–4 weeks of culture , the GFP-negative cell subset , expressing a quiescent effector memory cell ( TEM ) phenotype , is isolated by flow cytometry sorting . Approximately , 2–6% of the recovered cells carry latent HIV infection . Reactivation of virus replication is tracked by GFP expression , on an individual cell basis . The Spina model ( unpublished results; manuscript submitted ) is based on early work demonstrating that HIV-1 can establish infection directly in resting primary CD4+ T lymphocytes in vitro [26] , [27] , and on recent work showing that during acute HIV infection in a heterogeneous population of primary CD4+ T cells , undergoing varying degrees of cell activation , viral latency is established early and preferentially in non-dividing and minimally activated cells . This model uses the experimental approach of deriving latent NL4-3 infection ( wild-type ) in non-dividing “bystander” cells during brief co-culture with autologous productively infected , proliferating cells . When the quiescent bystander cell population is isolated from co-culture , the latently infected subset ranges from 1 to 12% cells containing integrated HIV DNA , and 0 . 5–5% cells with inducible provirus , as measured by expression of intracellular Gag . Latent infection is found in all of the major phenotypic subsets of T cells: naïve , central memory and effector memory . After incubations with experimental compounds , reactivation of virus replication is measured on a population basis , through quantification of tat mRNA by RT-qPCR . Verdin and colleagues have generated a number of Jurkat cell line-derived clones , bearing latent HIV-1 in single integration sites , that were engineered to express GFP in lieu of nef [8] ( J-Lat ) . J-Lat cells have been used in numerous studies that have contributed a wealth of knowledge in the area of viral latency . In contrast to several other models of HIV latency in cell lines , where mutations are present in the HIV tat gene or the TAR element , the J-Lat cell model contains wild-type tat and TAR . Three J-Lat clones established in the Verdin laboratory , 6 . 3 , 8 . 4 , 11 . 1 and one clone generated by the Greene laboratory , 5A8 , have been included in this comparison . J-Lat 5A8 was derived by specifically selecting for cells that would be more responsive to αCD3/αCD28 co-stimulation than the parental J-Lat line [28] . Under untreated basal conditions , little or no GFP expression is detected . However , reactivation of latent provirus is readily monitored by flow cytometry analysis of GFP expression . Results obtained with the above cell models were compared to results obtained in quantitative viral outgrowth assays ( QVOA; patient cell assay ) performed in the Margolis laboratory , with resting CD4+ T cells obtained from the leukopheresed peripheral blood of aviremic , ART-treated HIV-infected patients . This assay , as first described by three laboratories [1]–[3] , was later modified to its present design [22] . Following negative selection , resting CD4+ T cells are incubated with integrase and reverse transcriptase inhibitors to ensure the decay of any HIV genomes in the state of pre-integration latency [29] . The cells are exposed briefly to test compounds , and then plated in replicate microwells in a terminal-dilution assay and cultured with PHA-stimulated , allogeneic irradiated peripheral blood mononuclear cells ( PBMC ) from a sero-negative donor , and rIL-2 . After 19 days , the microcultures are scored for virus replication by soluble p24 production , and the number of cells containing replication-competent HIV is expressed as infectious units per million CD4+ T cells ( IUPM ) . Induction of viral reactivation across all cell models was assessed using a selected common panel of stimuli that are known to function by distinct and defined mechanisms of action . The panel included 13 treatments ( Table 2 ) that modulate T cell processes such as T-cell receptor engagement , protein kinase C ( PKC ) activation , calcium influx , cytokine signaling , histone deacetylation , and release of P-TEFb from the HEXIM/7SK RNP complex . This study was designed to answer the following questions: 1 ) are certain models of latency biased towards or against particular cell signaling pathways; 2 ) can stimuli be identified that work uniformly in multiple models; 3 ) can a central uniting theme or a single signaling pathway be responsible for control of viral latency; and 4 ) can a model or limited group of models predict experimental drug activity in authentic latently infected cells from patients ?
T-cell receptor engagement is effectively mimicked by the binding and cross-linking of antibodies against CD3ε , one of the signal transduction subunits in the CD3 complex [30] and the co-stimulatory molecule , CD28 [31] . Phytohemagglutinin ( PHA-M ) is a lectin that binds to carbohydrate moieties on surface glycoproteins . PHA is a polyclonal mitogen for T cells . Both PHA and αCD3/αCD28 antibody treatments stimulate signaling cascades that encompass TCR/LCK/p38 activation leading to calcineurin and NFAT activation , as well as PKC stimulation leading to NFκB activation . Incubation with αCD3/αCD28 antibody-coated beads produced strong responses in all primary cell models , with the exception of the Lewin model ( Figure 1A ) . In contrast , all J-Lat clones , except 5A8 were completely unresponsive to αCD3+αCD28 incubation ( Figure 1B ) . The response of J-Lat 5A8 cells after stimulation by αCD3+αCD28 coated beads , although detectable , was lower than that displayed by most primary cell models . However , the levels of stimulation can be improved using plate-bound αCD3 and free αCD28 antibodies , if so desired ( D . R . and W . C . G . , data not shown ) . Moreover , these cells are highly responsive to PHA , which indicates that these cells contain an intact signaling pathway downstream of TCR engagement . PHA reactivated latent viruses in all primary cell models and in the J-Lat clones , although with variable efficiency ( Figure 1 , Panels A and B ) . Therefore , the lack of responsiveness of J-Lat clones and of cells in the Lewin model to αCD3+αCD28 antibody treatment cannot be attributed to a lack of signaling mediators , since these cells respond to PHA through a highly similar signaling pathway . PKC is a family of ten kinases that are activated by phorbol esters [32] . In general , phorbol esters promote activation and differentiation of monocytes and monocytoid cells , as well as potent T-cell activation . Three PKC agonists were tested , namely PMA , prostratin ( both phorbol esters ) ; and bryostatin-1 ( a cyclic polyketide ) . PKC agonists activate the DAG-PKC-NFκB signaling pathway . PMA has long been used as a T-cell mitogen . PMA was tested at 2 nM in primary cell models and 16 nM in J-Lat clones . At these concentrations , PMA elicited maximal or near-maximal responses in J-Lat cells , except in clone 8 . 4 . Responses to PMA were near maximal in the Planelles and Siliciano models; the rest of the primary cell models and the QVOA also showed viral reactivation in response to PMA , although at more modest levels . Prostratin is a unique phorbol ester in that it induces potent T cell activation signals but , unlike PMA , is not tumorigenic . The ability of prostratin to induce T-cell activation through PKC , without tumor promoting ability , has made prostratin the subject of studies for its possible use as an inductive adjuvant therapy in the context of anti-retroviral therapy ( ART ) [33] . Another unique property of prostratin is that , despite being able to reactivate latent HIV-1 , it exerts an inhibitory effect on active HIV-1 replication through downregulation of CD4 [34] , [35] . The relative reactivation efficiencies observed in response to prostratin were similar to those obtained with PMA treatment . Thus , the models with the highest responses to PMA ( J-Lat 6 . 3 and 11 . 1 clones , and Siliciano and Planelles models ) showed the highest responses to prostratin as well . Conversely , poor to intermediate responses to PMA , observed in the Greene , Lewin and Spina models , and the quantitative patient-cell outgrowth assay ( QVOA ) were paralleled by similar responses to prostratin ( Figure 1 , Panels A and B ) . In the specific case of the Greene model , it has been observed that only a minority of cells , about 5% , respond to PMA , although the reasons for this observation are unknown . Bryostatins are a family of natural products found in several species of bryozoans . Bacterial symbionts of the bryozoan species are thought to be responsible for bryostatin synthesis ( reviewed in [36] ) . Bryostatins bind to the diacylglycerol-binding region within the C-1 regulatory domain of PKC . Bryostatin-1 was recently shown to reactivate latent HIV-1 in vitro in monocytoid and lymphoid cell line models of latency [37] and was approximately 1 , 000-fold more potent than prostratin . More recently , DeChristopher and colleagues achieved the chemical synthesis of several analogs of bryostatin-1 , which demonstrated potent activity in J-Lat cells [38] . Bryostatin-1 was very potent in J-Lat clone 11 . 1 , but had only modest activity in the other J-Lat clones ( Figure 1B ) . In primary cell models , bryostatin-1 induced maximal response in the Siliciano model , and about half-maximal responses in the Lewin and Planelles models . However , the Greene and Spina models , and the patient cell outgrowth assay , showed very low to non-detectable responses to bryostatin-1 ( Figure 1A ) . A commonly utilized T-cell activation regimen in the laboratory , which mimics the signaling pathway used in TCR engagement , is the combination of PKC activation via PMA along with the calcium ionophore , Ionomycin , which bypasses the requirement for both CD3/TCR and CD28 receptor engagements . Signaling downstream of TCR engagement involves the formation of inositol triphosphate , which triggers an increase in the intracellular Ca2+ concentrations , which in turn activate the phosphatase , calcineurin . Calcineurin then dephosphorylates cytoplasmic NFAT transcription factor , which translocates to the nucleus . A combination of PMA and Ionomycin induced vigorous viral reactivation in most cell models tested , but not in the Spina model . Viral reactivation in response to PMA+Ionomycin was generally increased when compared to that of PMA alone , with the exception of the Lewin and Spina models ( Figure 1 , panels A and B ) . Unexpectedly , PMA+Ionomycin stimulation of primary T-cells in the Spina model caused inhibition of Tat mRNA transcription , the readout in this assay , to below initial basal levels ( Figure 1A ) . It has been reported previously that PMA induction of HIV replication can be Tat-independent [39]; and in this case , the combination with Ionomycin appeared to actually suppress Tat transcription at 24 hrs . following stimulation . In the patient cell outgrowth assay/QVOA , PMA+Ionomycin produced a strong reactivation response that was higher than that observed with each compound alone . Previous reports showed that incubation with IL-7 , alone [40] or in combination with IL-2 [41] can reactivate latent HIV-1 in resting CD4+ T cells isolated from infected individuals . IL-7 also reactivated latent HIV-1 in thymocytes in a SCID-hu mouse model of HIV latency [42] and in cultured TCM in the Planelles model [43] . In the Planelles model , IL-2+IL-7 stimulation of latently infected cells was previously shown to be inefficient ( 10–20% of the reactivation obtained with αCD3/αCD28 ) and to promote division of infected cells in the absence of viral reactivation [43] . Responsiveness to IL-7 , or IL-2+IL-7 stimulation is physiologically relevant as these cytokines , along with IL-15 , are known to drive the homeostatic proliferation of memory T cells in vivo [44] . A recent study found that IL-7 , when administered to HIV-1 infected patients undergoing ART , promotes viral persistence by enhancing residual levels of viral production and inducing proliferation of latently infected cells without reactivation [45] . Robust responsiveness to IL-2+IL-7 was observed in the Siliciano and Spina primary cell models , and minimal activity was observed in the Greene model . Cells in the Lewin and Planelles models and the patient cell outgrowth assay responded poorly or not at all ( <5% of maximal ) ; whereas , cells in the Greene model exhibited a weak response . It is interesting to note that IL-7 used alone at 25 ng/ml induced robust reactivation in the Lewin model [46] . J-Lat cells failed to reactivate virus in response to IL-2+IL-7 stimulation . Jurkat cells , the parental tumor cell line from which J-Lat clones were derived , are IL-2-independent for their growth and survival , do not express the high-affinity IL-2 receptor , CD25 [47] , [48] , and express low levels of the IL-7 receptor alpha [49] . TNF-α is a potent inducer of viral gene expression in certain tumor cell lines harboring integrated , latent HIV-1 , through the activation of NFκB [5] , [8] , [50] , [51] . As previously reported , TNF-α treatment activated virus expression in J-Lat cells , especially in clones 6 . 3 and 11 . 1 ( Figure 1B ) . However , among the primary cell models , TNF-α failed to induce any detectable viral reactivation in the Greene and Planelles models and showed only minimal activity in the Lewin and Siliciano models . In contrast , the patient cell outgrowth assay responded robustly to TNF-α , and cells in the Spina model showed an intermediate response . In order to better understand the responsiveness , or lack thereof , to TNF-α , we analyzed the levels of TNF-R in primary and Jurkat cells . We isolated bulk PBMC from two donors , selected memory CD4+ cells using CD45RO expression , and then stained the cells for CCR7 , CD27 and the TNF-α receptor . These experiments showed that none of the freshly selected memory subsets tested ( specifically , TCM , TEM and transitional memory T cells , TTM ) expressed detectable levels of the TNF-α receptor ( Figure S2 ) . TNF-R expression was extremely low in cultured TCM from the Planelles model ( Figure S2 ) . In contrast , J-Lat 10 . 6 cells expressed high levels of TNF-R ( Figure S2 ) . HIV reactivation in response to TNF-α in vitro and in vivo is likely linked to whether cells under the specific culture or physiological conditions upregulate the expression of the TNF-α receptor . Hexamethylene bisacetamide ( HMBA ) is a hybrid bipolar compound that induces differentiation and apoptosis in transformed cell lines in culture [52] , [53] . HMBA was shown to activate HIV transcription in vitro [7] , [54] , to reactivate latent HIV in vitro [19] , [55] and to reactivate HIV in primary cells from aviremic , infected patients [56] . The activity of HMBA on HIV transcription is a result of its ability to induce dissociation of P-TEFb from the inhibitory 7SK ribonucleoprotein complex [19] , [55] . HMBA treatment had significant reactivation activity in the QVOA and the Lewin model , but demonstrated little to no activity in the rest of the primary cell models and J-Lat clones tested ( Figure 1 , panels A and B ) . The “histone code” model states that a variety of covalent , post-translational modifications ( PTM ) on histone tail residues regulate the interaction of transcriptional regulators with chromatin to determine gene expression levels . The nature and localization of such post-translational modifications is broad , and their ability to act in a combinatorial manner provides an attractive model for how a finely tuned regulation can be effected . Histone code modifications include acetylation , phosphorylation , methylation , ubiquitination and sumoylation , among others [57] , [58] . Acetylation of lysine residues in histone tails can have two important effects on chromatin organization ( reviewed in [58] ) . First , this PTM results in neutralization of a basic charge on the lysine residue , which results in disruption of histone contacts with other histones and with DNA , diminishing the degree of compaction of the local chromatin . Second , proteins containing a specialized domain known as bromodomain specifically recognize acetylated lysine residues and then trigger downstream regulatory effects . Acetylation of histones is regulated by the concerted action of HATs and HDACs . Acetylated histones have long been associated with actively transcribed genes [59] and , therefore , inhibitors of HDAC ( HDACi ) are considered as general activators of transcription . Two main categories of HDACs have been described: Class I ( HDAC 1 , 2 , 3 and 8 ) , and Class II ( HDAC 4 , 5 , 6 , 7 , 9 , 10 and 11 ) . Inhibition of Class I , but not Class II , HDACs has been shown to induce reactivation of latent HIV [60] , [61] . Suberoylanilide hydroxamic acid ( SAHA; also known as vorinostat ) is a pan-HDAC inhibitor that targets both Class I and Class II HDACs , and can induce reactivation of HIV in models of HIV latency [11] , [62]–[65] , and in resting cells from ART-treated , aviremic HIV-infected patients [65]–[67] , although it failed to induce reactivation in patient cells in another study [68] . Recently , a single administration of SAHA to ART-treated , aviremic patients was shown to induce global cellular acetylation and increases in viral RNA in resting CD4+ cells from these patients [69] . To test the ability of HDAC inhibitors to reactivate latent HIV in the various models of latency , we utilized three such inhibitors , provided by Merck Research Laboratories . SAHA potently blocks the Class I HDACs ( i . e . , 1 , 2 , 3 , and 8 ) and has modest activity against Class II HDACs ( i . e . , 6 , 10 and 11 ) . MRK-1 is a selective inhibitor of the Class I HDACs ( i . e . , 1 , 2 and 3 ) and HDAC6 ( Class II ) ; whereas , MRK-11 selectively blocks Class II HDACs ( i . e . , 4 , 5 , 6 and 7 ) and HDAC8 ( Class I ) [60] . SAHA was moderately potent in the Lewin and Spina cell models and QVOA , but was marginally active or inactive in the rest of the primary cell models and the J-Lat clones . The activity profile of MRK-1 was similar to that of SAHA in the primary models , showing the best activity in the Lewin cell model and the patient cell outgrowth assay . All the J-Lat clones had modest responses to MRK-1 , which contrasted with the poor activity seen with SAHA in these cells ( Figure 1B ) . The differences between SAHA and MRK-1 responses could , potentially be explained by the slightly different specificities of these HDAC inhibitors . In general , MRK-11 was inactive or minimally active ( <20% response ) in the QVOA and all J-Lat and primary cell models , except in the Lewin model , where it exhibited close to 50% activity . Cells in the Lewin model are unique in this study , in that they are very sensitive to viral reactivation by both Class I and Class II HDAC inhibitors . In contrast , other models tested are either insensitive to HDACi or show sensitivity to Class I inhibitors but not to Class II . The relationship between models based on the ability of compounds to activate latent HIV within each model was investigated by hierarchical clustering and heatmap visualization ( Figures 2A and 2B ) . Two comparisons were performed . First , all the cell models for which data was available for all compounds and at all concentrations were compared ( Figure 2A ) . This comparison excluded the patient cell outgrowth assay for which data for only certain concentrations of activators were available . In the second comparison , all models were included but only those concentrations that were universally tested were included ( Figure 2B ) . In both comparisons , reactivation values obtained with PHA at 10 µg/ml were used as a reference , to which all other reactivation values were normalized to . Both comparisons yielded strikingly similar results . Three significant clusters of models were identified , with one robust outlier , the Spina model . The Lewin and J-Lat 5A8 clustered very close in both comparisons ( Figures 2A and 2B ) , with the patient cell assay/QVOA being the next closest to those two ( Figure 2B ) . Therefore , the first subcluster is defined by the Lewin , J-Lat-5A8 and QVOA models . The second subcluster is defined by the Planelles and Siliciano models , closest to each other , and the Greene model . The first two subclusters have a close association with each other , that separates them from the three remaining J-Lat clones ( 8 . 4 , 6 . 3 and 11 . 1 ) , which form the third and more distant subcluster . This clustering conforms to what would be expected biologically with the majority of primary cell models clustering together and the majority of cell line models clustering separately , with the exception of J-Lat 5A8 , which clusters among the primary models . In addition , this clustering pattern was largely maintained when the QVOA data was included and a reduced compound set analyzed ( Figure 2B ) . Since all primary cell models clustered together , this suggests that the resting phenotype of these models compared with the proliferating phenotype of J-Lat cells may influence the responsiveness to different agents . The QVOA model appears to cluster robustly with the Lewin model and the J-Lat 5A8 , suggesting that these two models may represent the best proxy currently available for the activation capabilities of compounds when analyzing cells from HIV-infected subjects . However , this interpretation should be treated with caution as the clustering in Figure 2B , when the QVOA data was included , was performed with a reduced compound set and may not be as robust as the analysis that included all compounds at all concentrations ( Figure 2A ) . The relationship of compounds to each other , based on their ability to activate HIV across the different models , was also investigated by hierarchical clustering and heatmap visualization ( Figures 2A and 2B ) . The first analysis ( Figure 2A ) revealed that PMA+Ionomycin and , separately , αCD3+αCD28 antibody stimulation represented treatments that were strong outliers . The rest of the compounds then fell into one of two significant major clusters . The first cluster contained the majority of the HDACi , but also IL-7+IL-2 treatment , Ionomycin , and HMBA . The second cluster contained all concentrations of the PKC activators ( i . e . , prostratin , PMA and bryostatin ) as well as PHA , TNF-α and the 6 µM concentration of MRK-1 . This pattern of compound clustering was supported when data from the QVOA was included and a reduced compound set analyzed ( Figure 2B ) . It is noteworthy that HMBA clustered interspersed with the HDAC inhibitors , which suggests potential similarities in the mechanism of action . The recent finding that the HDAC inhibitor , SAHA , can release P-TEFb from the inhibitory 7SK snRNP complex [70] provides a potential explanation for the close clustering of HMBA and HDAC inhibitors . In fact , a provocative finding in that study was that the viral reactivating ability of SAHA did not correlate with histone H3 or tubulin acetylation but , rather , with release of P-TEFb [70] . As shown in Figures 2A and 2B , the NFκB agonists PMA , prostratin , bryostatin , PHA and TNF-α cluster together . This result indicates that NFκB agonists consistently work as latency-reversing drugs across the different models , and that NFκB may play a central role in viral reactivation from latency , independent of the model used . In agreement with that , PHA and PMA were active in all the models tested . In summary , the clustering of compounds based on their activation of HIV across models conforms to what would be expected biologically and validates the analytical approach utilized in the current study .
This study represents the first experimental comparison among several broadly used HIV latency systems , including primary cell models , transformed cell lines and patient-derived cells . To establish these comparisons in an unbiased manner , we chose a panel of known stimuli that were tested in parallel in the selected cell models . The methodology was designed to circumvent variations due to batch , formulation or concentration differences in the compounds tested . To the extent possible , the duration of exposure to each stimulus , the inclusion of appropriate controls and the maximal-response stimulus were standardized as well . PHA was the only stimulus that uniformly reactivated latent viruses in all systems tested . Most T cells , whether transformed or primary , express CD3ε or CD2 , both of which are triggered by PHA . Unfortunately , the therapeutic potential of agonists of the CD3/CD28/CD2 signaling pathway is uncertain , given the plethora of undesirable side effects , including transient lymphopenia , previously observed in patients treated with OKT3 antibodies [71] , [72] . PMA also reactivated viruses across models . Responsiveness to PMA was roughly , although not exactly , paralleled by responsiveness to the other PKC agonists tested , prostratin and bryostatin . For example , patient cells were responsive to PMA and prostratin , but not to bryostatin . Differences may be explained by the repertoire of PKC isoforms that is activated by each PKC agonist . This issue will require further exploration , as it is likely that certain PKC isoforms may be more involved than others in the reactivation of latent HIV . It is also plausible that certain PKC isoforms may be able to mediate viral activation with only minimal induction of cellular activation and/or proliferation , which , if true , would clearly be desirable in an eradication strategy . The addition of Ionomycin to PMA generally provided an enhancement of the activity observed with PMA alone , with the exception of cells in the Spina model . This is intriguing , and contrary to expectations . Ionomycin induces calcium influx , which activates the calcineurin phosphatase that , in turn , activates NFAT . A possible explanation for the loss of activity with PMA+Ionomycin in the Spina model might be the onset of apoptosis , due to a high level of stimulation . However , this was not the case; increased cell death was not observed in these cultures during testing . Virus reactivation in the Spina model was measured by levels of tat mRNA transcription after 24 hrs . following exposure to stimulus . In other studies , in which HIV reactivation was tracked by production of soluble p24 , virus replication was detected readily 4–5 days after PMA+Ionomycin stimulation ( C . A . S . , unpublished results ) . Because PMA+Ionomycin stimulation delivers such strong and immediate cell activation signals , it may be possible that at early time points , limited “signaling resources” in primary T cells could be redirected away from the viral LTR and initiation of tat transcription [39] . Additional studies will be necessary to address this mechanistic point . The activities of cytokines are usually dependent on the presence or absence of their respective receptors on the target cells . TNF-α showed remarkable activity in several J-Lat clones and in patient cells , but was inactive or had low ( Lewin ) to moderate ( Spina ) activity in the primary cell models . As stated above , the TNF-R was not found in cultured or fresh TCM . Therefore , the high level of responsiveness in patient cells may underlie upregulation of the receptor under the culture conditions utilized , including perhaps the incubation with TNF-α itself . It will be informative to ascertain whether such upregulation occurs , and the specific conditions influencing it . This putative upregulation of the TNF-R is potentially exciting because , if appropriately targeted to cells in the latent reservoir , it would render cells exquisitely responsive to TNF-α or an agonist thereof . Responsiveness to TNF-α clusters among PKC agonists ( Figures 2A and 2B ) , which likely reflects the fact that both types of stimuli culminate in NFκB activation . However , in the analysis displayed in Figure 2A , TNF-α clusters closest with MRK-1 , an HDACi , at 6 µM . HDAC inhibitors are the first drug class to be utilized in clinical trials for HIV eradication and the results so far have been promising [69] because intracellular increases in HIV transcription were induced in vivo during SAHA treatment . Future development of HDAC inhibitors should be directed at ascertaining which HDAC isoforms are more involved in maintaining HIV latency , so that they can be specifically targeted . In general , the Lewin model clustered closely with the J-Lat 5A8 cells and both of these clustered with the patient cell outgrowth assay . However , one of the major differences between both models pertains to responsiveness to the HDACi , MRK11 , which blocks Class II enzymes . Cells in the Lewin model displayed very high sensitivity to all tested HDACi , and were the only ones in this study to exhibit a substantial response to MRK11 . In contrast , patient cells in the outgrowth assay did not respond to MRK 11 . Three primary cell models , Greene , Planelles and Siliciano , had extremely low or no sensitivity to HDACi treatments . It is unclear what aspects of the biology of the cells or the latent viruses in these models renders the latent viruses so refractory to the effects of HDAC inhibition . As we suggest below , the low levels of active P-TEFb components in resting cells may constitute a major barrier to efficient transcription , which may not be overcome simply by inhibition of HDACs . Recent observations indicate that incubation of primary resting cells with stimuli that induce P-TEFb allows the cells to then become responsive to HDAC inhibition ( Matija Peterlin , UCSF; personal communication ) . Cells in the Lewin and patient cell/QVOA models shared responsiveness to HMBA , while most other models had very low or no responsiveness to this agent . HMBA facilitates the dissociation of P-TEFb from the 7SK snRNP complex and makes P-TEFb more readily available to interact with Tat , and then to be recruited to the TAR loop on nascent viral RNAs . This is an early step in the transcriptional activation of the silent provirus and , therefore , it is viewed as a “gate keeper” step . Recent reports [18] , [70] , [73] have suggested that resting T cells contain very low levels of cyclin T and phosphorylated CDK9 , leading to the hypothesis that the activity of the P-TEFb complex is inherently low , and not controlled by recruitment to the inactive 7SK snRNP complex . In view of these observations , Budhiraja et al . explained the lack of responsiveness of cells in the Planelles model as a result of the low levels of cyclin T and of CDK9 phosphorylation [73] . However , the previous model does not explain two of the observed responses in the present studies . First , patient cells and those in the Lewin model responded strongly to HMBA , while also being quiescent . Future studies should be undertaken to test the levels of P-TEFb in these model systems and examine the correlation between levels of P-TEFb and sensitivity to HMBA . Second , J-Lat cells seemed unresponsive to HMBA , while they would be expected to have high levels of active P-TEFb , given that they are dividing cells . We speculate that P-TEFb is not limiting in J-Lat cells , and that the rate-limiting step to active proviral transcription is either at the transcription initiation level , prior to the participation of Tat , or downstream of P-TEFb recruitment . A plausible mechanism for the lack of activity of HMBA in J-Lat cells is through transcriptional interference imposed by a proximal cellular promoter , as was shown for certain J-Lat clones , including J-Lat 6 . 3 and 8 . 4 [74] . Ionomycin was a poor inducer of reactivation in all primary cell models and patient cells , and had no detectable activity in the J-Lat cells . Calcium influx is necessary for activation of the NFAT transcription factor , but is not sufficient by itself for optimal viral reactivation . It appears that the full effect of NFAT on HIV reactivation , at least in cultured TCM , requires an additional signal provided by LCK activation [15] . PKC agonists were generally potent reactivators in most models tested here . Bryostatin is of particular interest because it stands as the only PKC agonist that is FDA approved and , consequently , data on its pharmacokinetics and toxicity in humans are available [75] , [76] . Bryostatin has been tested in clinical trials for cancer and Alzheimer's disease [75] , [76] . In addition , bryostatin was shown to synergize with the HDACi , valproic acid , in reactivation of latent HIV in a J-Lat model [76] . Although bryostatins are emerging as potential therapeutics for HIV eradication , they typically induce cellular activation , proliferation and secretion of pro-inflammatory cytokines . Thus , future research will need to identify analogs with diminished capacity to induce such undesirable cellular effects , while preserving the ability to reactivate latent HIV . No single experimental system of HIV latency completely recapitulated responsiveness to all types of stimuli tested here . The Lewin in vitro model displayed the broadest responsiveness . Similarities between responsiveness of patient-derived cells and the Lewin model cells were observed more frequently than with any other model . However , several notable differences separated the previous two models . These were: the high responsiveness of the Lewin model to MRK-11 and bryostatin , which contrasted with the lack of responsiveness of patient cells; and the lack of response of Lewin cells to αCD3/αCD28 . The lack of responsiveness of the Lewin model cells to IL-2+IL-7 contrasted with the high responsiveness of the Spina and Siliciano models . Therefore , secondary screening of latency reversing drugs obtained through high throughput systems could be accomplished by using a combination of testing in the Lewin system plus a system that shows complementary properties , such as the Spina or the Siliciano models . The site of proviral integration can modulate the levels of viral transcription and has been proposed as a mechanism to explain latency [77] , [78] . Specifically , integration in the vicinity of actively transcribed cellular genes can lead to transcriptional interference effects [74] , [79] , [80] . The present study did not attempt to analyze the influence of integration on proviral latency status . However , in a separate study [81] , the influence of host cell gene transcription on proviral latency was analyzed and compared for five different models of latency including the Siliciano [17] and Planelles [15] models , a Jurkat model with polyclonal integration [78] , infection of primary resting CD4+ T cells [82] , and infection of primary activated CD4+ T cells [82] . When the influence of positioning in the chromosome ( regardless of orientation ) was examined , proviruses integrated in nearby positions shared the same latency status more often than predicted by chance . However , this trend was only statistically significant when comparing proviruses within each model , but not when comparing proviruses across models . This was interpreted by the authors to mean that local chromosomal features affecting latency are model-specific . Regarding proviral orientation with respect to cellular genes , the Siliciano model exhibited a modest , but statistically significant preference for latent proviruses to be in the same orientation as proximal cellular genes , confirming a previous report [80] . In contrast , the other models exhibited no statistically significant deviation from 50% of latent integrations being in the same orientation as cellular genes . Rational design of drugs to target HIV latency is not possible at the moment , because we do not have precise knowledge of all the cellular factors and activation pathways that impact viral transcription , leading to productive replication . A second obstacle to rational drug design for viral eradication lies in the notion that while the desired compound should trigger HIV reactivation , it should induce minimal or no cellular activation/proliferation . Therefore , drug screening studies should include an evaluation of the ability of candidate compounds to induce expression of cellular activation markers and proliferation .
Studies involving human peripheral blood mononuclear cells were conducted at the following institutions , and approved by the respective internal boards as indicated: The test compounds , listed in Table 2 , were obtained , and stocks prepared and distributed centrally to each of the participating laboratories by the CARE Pharmacology Core of the University of North Carolina . The compounds were tested in each cell model at the following final concentrations: αCD3/αCD28-conjugated beads ( Dynal ) at 1∶1 bead∶cell ratio; PHA-M ( Sigma ) at 1 . 1 , 3 . 3 , 10 µg/mL; PMA ( Sigma ) at 2 nM for primary T cells , 16 nM for J-Lat cells; Ionomycin ( Sigma ) at 0 . 5 µM; prostratin ( LC Laboratories ) at 0 . 3 , 1 , 3 µM; bryostatin ( provided by the National Cancer Institute ) at 10 , 33 , 100 nM; SAHA/vorinostat ( Merck ) at 0 . 11 , 0 . 33 , 1 µM; MRK-1 ( class I HDACi , Merck ) at 0 . 67 , 2 , 6 µM; MRK-11 ( class II HDACi , Merck ) at 3 , 10 , 30 µM; HMBA ( Sigma ) at 0 . 3 , 1 , 3 mM; TNF-α ( Peprotech ) at 10 ng/mL; IL-2 ( Peprotech ) at 30 IU/mL; IL-7 ( Peprotech ) at 50 ng/ml . IL-2 , IL-7 , TNF-α , and αCD3/αCD28 bead stocks were prepared in RPMI culture medium; HMBA stock was prepared in water . All the other compounds were prepared in DMSO solvent . Unless otherwise specified , each cell model tested and the HIV outgrowth assay included the controls: untreated ( base culture medium ) , 0 . 1% DMSO , 0 . 5% DMSO ( specific to 10 µg/mL PHA ) . The exposure time of cells to compounds was standardized across the models to 24 hrs . , except for PHA ( 48 hrs . ) , αCD3/αCD28 beads ( 48–72 hrs . ) , and IL-2+IL-7 ( 5 days ) . The timing of assay read-outs for HIV reactivation was specific to each model system , dependent on unique cellular and viral characteristics . Initially any compound or any concentration of a compound that was not used universally across all models was removed . The untreated control , representing background activation , was subtracted from each compound for each donor in each model . Activation values for each compound were then averaged across donors within each model and any activation resulting from the DMSO condition was subtracted from those compounds that were dissolved in DMSO . DMSO has structural similarity to HDAC inhibitors , as some of these compounds were derived from DMSO following the observation of DMSO effects on transformed cells [85] . Average activation values for each compound were then normalized within each model by dividing by the average activation value for the highest concentration of PHA used so that models could be compared to each other . Finally , examining the distribution of average activation values across compounds revealed right-skewed data for each model and thus a log10 transformation was performed . Constants were added to the average activation value for each compound to account for negative values prior to log10 transformation and to shift activation values into a range that reflected their actual activation level . An unsupervised approach was used to determine the relationship between compounds based on their ability to activate HIV across models and also between models based on their response to compounds . Cluster 3 . 0 [86] was used for hierarchical clustering of compounds and models such that distances were calculated using the Euclidean based metric and then clustered using the average linkage method . The results were visualized in a heatmap using Java TreeView [87] . The statistical significance associated with clustering was determined using pvclust [88] ( R package ) , which calculates approximately unbiased ( AU ) p-values that are computed using multiscale bootstrap resampling such that 95% equates to a p-value cut-off of 0 . 05 . These normalization procedures and hierarchical clustering approaches were performed twice since not every compound was assessed at every concentration in the QVOA model . Specifically , the were performed once using a complete list of compounds but without data from the QVOA and a second time with a subset of compounds but now with the inclusion of data from the QVOA . TNFR surface expression was determined using anti-human TNFRI-APC ( R&D Systems , Minneapolis , MN ) . Briefly , 1×105 cells were incubated with 1∶100 anti-human TNFRI-APC in 100 µl of PBS/3%FBS Buffer during 30 min at 4°C followed by flow cytometric analysis in a BD FacsCanto II flow cytometer using the FACSDiva software ( Becton Dickinson , Mountain View , CA ) . Data was analyzed with FlowJo ( TreeStar Inc . , Ashland , OR ) .
|
HIV establishes a state of latency in vivo and this latent reservoir , although small , is difficult to eradicate . To be able to better understand this state of latency , and to develop strategies to eliminate it , many groups have developed in vitro models of HIV latency . However , notable differences exist among cell model systems because compounds that reactivate latent HIV in a particular system often fail to do so uniformly across different models . To begin to understand the biological characteristics that are inherent to each HIV model of latency , we compared the response properties of five primary T cell , four J-Lat cell models and those obtained with patient-derived infected cells . A panel of thirteen stimuli that are known to reactivate HIV by defined mechanisms of action was selected and tested in parallel in all models .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2013
|
An In-Depth Comparison of Latent HIV-1 Reactivation in Multiple Cell Model Systems and Resting CD4+ T Cells from Aviremic Patients
|
Viruses that naturally infect cells expressing both MHC I and MHC II molecules render themselves potentially visible to both CD8+ and CD4+ T cells through the de novo expression of viral antigens . Here we use one such pathogen , the B-lymphotropic Epstein-Barr virus ( EBV ) , to examine the kinetics of these processes in the virally-infected cell , comparing newly synthesised polypeptides versus the mature protein pool as viral antigen sources for MHC I- and MHC II-restricted presentation . EBV-transformed B cell lines were established in which the expression of two cognate EBV antigens , EBNA1 and EBNA3B , could be induced and then completely suppressed by doxycycline-regulation . These cells were used as targets for CD8+ and CD4+ T cell clones to a range of EBNA1 and EBNA3B epitopes . For both antigens , when synthesis was induced , CD8 epitope display rose quickly to near maximum within 24 h , well before steady state levels of mature protein had been reached , whereas CD4 epitope presentation was delayed by 36–48 h and rose only slowly thereafter . When antigen expression was suppressed , despite the persistence of mature protein , CD8 epitope display fell rapidly at rates similar to that seen for the MHC I/epitope half-life in peptide pulse-chase experiments . By contrast , CD4 epitope display persisted for many days and , following peptide stripping , recovered well on cells in the absence of new antigen synthesis . We infer that , in virally-infected MHC I/II-positive cells , newly-synthesised polypeptides are the dominant source of antigen feeding the MHC I pathway , whereas the MHC II pathway is fed by the mature protein pool . Hence , newly-infected cells are rapidly visible only to the CD8 response; by contrast , latent infections , in which viral gene expression has been extinguished yet viral proteins persist , will remain visible to CD4+ T cells .
Many intracellular pathogens , particularly viruses , naturally infect cells of the haemopoietic system that express both MHC I and MHC II molecules . Such infected cells may be rendered visible to the host T cell response through the intracellular processing of virally-encoded proteins , leading to cell surface display of MHC I- and MHC II- peptide complexes recognised by CD8+ and CD4+ T cells respectively . With regard to MHC I-restricted presentation , the speed with which virus-infected cells become recognisable by CD8+ T cells [1] and the involvement of the proteasome in that process [2] led to the idea that a proportion of all newly-synthesised viral polypeptides were marked for immediate degradation , generating peptides that were fed into the MHC I pathway [3] . While the concept has evidential support [4] , [5] , [6] , [7] , questions remain about the proportion of translation products sacrificed in this way [8] , [9] , the mechanism that underpins their selection [10] , [11] and most importantly the degree to which , in latently-infected cells where viral antigen synthesis has been extinguished , cells may still be visible to the virus-specific CD8 response through MHC I-restricted processing of antigen from the mature protein pool . Only two studies have attempted to address this latter issue by specifically regulating antigen expression rather than resorting to general inhibitors of translation [12] , [13] . Though both studies supported the dominance of newly-synthesised protein as an antigen source , in each case the evidence came from a single epitope studied at a very limited number of time points leaving the generality of the results , with respect to such variables as antigen dose , epitope location and target cell identity , unresolved . Less is known about the rules governing MHC II-restricted presentation of endogenously expressed viral antigens , though it is clear that under some circumstances this can occur [14] , [15] . To date there are examples of endogenous antigen accessing the MHC II pathway either through location in the endoplasmic reticulum itself [16] , through delivery to endosomes/lysosomes by macro- [17] , [18] or chaperone-mediated [19] autophagy , or through release and re-uptake by neighbouring cells [20] . However there is little information on two important issues: firstly the kinetics with which MHC II-restricted epitopes are presented following antigen expression , which determines when a newly-infected cell becomes visible to the CD4+ T cell response , and secondly the relative importance of newly-synthesised polypeptides and the mature protein pool as antigen sources . Here we address these issues using Epstein-Barr virus ( EBV ) , a human gamma-herpesvirus that transforms B cells in vitro into MHC I/II-positive lymphoblastoid cell lines ( LCLs ) expressing eight viral proteins , the nuclear antigens EBNAs 1 , 2 3A , 3B , 3C and -LP , and the latent membrane proteins LMPs 1 and 2 [21] . Such LCLs resemble the virus-transformed B cells that arise during EBV infection in vivo and elicit the MHC I- and MHC II-restricted T cell responses that control the infection [22] . Many of these responses have been mapped to individual peptide epitopes and epitope-specific CD4+ and CD8+ T cell clones shown to recognise MHC-matched LCL targets [22] . Here we sought to use such clones to follow the presentation of EBV antigens via the MHC I and MHC II pathways in an LCL background which lacked base-line epitope display and where expression of the cognate antigen could be temporally controlled . For this purpose we chose two indicator antigens , EBNA3B and EBNA1 . EBNA3B is non-essential for transformation in vitro and therefore one can establish LCLs with an EBNA3B gene-deleted virus [23] , [24]; EBNA1 , the virus genome maintenance protein , is required for transformation but shows sequence variation between virus isolates , allowing one to establish LCLs using a virus that lacks many of the relevant T cell epitopes [25] , [26] . In both cases we then introduced the cognate antigen-coding sequence under the control of a doxycycline-regulated promoter and monitored CD4 and CD8 epitope display after inducing or suppressing new antigen synthesis .
Figure 1A shows the vector used to achieve dox-dependent antigen expression [27] . Rat CD2 expression from the vector backbone allows initial enrichment of transfected cells , while the EBV ori-p sequence promotes episomal maintenance in LCLs . Antigen-coding sequences lie under the control of a dox-regulated promoter . We first introduced an EBNA3B-carrying vector ( pEBNA3B-tet , Figure 1A ) into LCLs made using a recombinant EBNA3B-KO virus . Figure 1B illustrates the pattern of results consistently observed with stable pEBNA3B-tet transfectants on three different LCL backgrounds . EBNA3B protein expression , undetectable by immunoblotting in non-induced cells , showed a clear dose-dependent response to 7 day treatment with dox , reaching a level equivalent to that seen in wild-type EBV-transformed LCLs at 25 ng/ml dox and increasing to supra-physiologic levels at higher dox concentrations . We then assayed these same cells , dox-induced for 7 days , as targets for T cell recognition . CD8+ T cell clones were generated against five well-defined epitopes in EBNA3B ( HRC/B*2705 , RRA/B*2702; AVF/A*1101 , IVT/A*1101 , VEI/B*4402; positions shown in Figure 1A , see Table S1 for details ) . Because EBNA3B had not been studied before as a CD4 target , we first screened EBV-immune donors for CD4+ T cell reactivity to an EBNA3B peptide panel in IFNγ Elispot assays , generated CD4+ T cell clones against three of the epitopes thus defined and determined their MHC II restriction using standard approaches [28] , [29] . These epitopes ( FIE/DRB1*1501 , ILR/DRB4*01 and QAP/DRB3*0201; see Table S1 ) are located on the EBNA3B sequence in Figure 1A . Figure 1C shows representative results from such experiments , here using a pEBNA3B-tet LCL ( A*1101/DRB3*0201-positive ) as a target for CD8+ clones against the IVT/A*1101 epitope and for CD4+ clones against the QAP/DRB3*0201 epitope . All such experiments included , as a positive control target , a wild-type EBV-transformed LCL from the same individual expressing EBNA3B from the resident EBV genome . Target cell recognition is assayed by IFNγ release after 18 h of co-culture . There was no response to the non-induced pEBNA3B-tet LCL by either CD8+ or CD4+ effectors , whereas dox-induced cells were recognised at levels which increased in a dose-dependent manner . For both effector populations , the recognition of target cells exposed to 25 ng/ml dox ( i . e . the dose inducing physiologic levels of EBNA3B ) was similar to that seen for the wild-type LCL , whereas higher levels of induction increased recognition accordingly . Assays with different pEBNA3B-tet LCLs , using effector cells against the other four CD8 and two CD4 epitopes in EBNA3B , gave very similar results ( data not shown ) . All subsequent studies were therefore conducted on cells induced to express indicator antigens at physiologic ( 25 ng/ml dox ) and at supra-physiologic ( 100 ng/ml dox ) levels , with similar patterns of results obtained . We first asked how quickly target cells became susceptible to CD8+ and CD4+ T cell recognition following dox-induction . Figure 2 shows one such experiment inducing the above pEBNA3B-tet LCL at the two dox concentrations . In both cases , expression of EBNA3B protein was detectable by immunoblotting within 6 h of dox addition , and by 72 h had increased to reach a stable steady-state level that was again higher ( relative to a wild-type LCL ) at the higher inducing dose ( Figure 2A ) . Aliquots of the same cells were used as targets in T cell assays , each time alongside cells from the appropriate non-induced and long-term-induced cultures . To examine epitope display at the precise time of harvest , all target cells were fixed in 1% PFA before addition to the assay . As shown in Figure 2B , while absolute levels of IFNγ release were always higher with targets given 100 ng/ml dox , the same pattern of results was obtained following antigen induction at either dose . Thus , recognition by CD8+ T cells specific for the IVT/A*1101 epitope was detectable within 6 h of dox induction and by 36 h had increased to plateau at the same level as seen against long-term dox-induced targets . In contrast , recognition by CD4+ T cells specific for the ILR/DRB4*01 epitope was not detectable until 36–48 h and increased quite slowly thereafter , only reaching the long-term dox plateau level on targets induced for 168 h . In further experiments with this and other pEBNA3B-tet LCLs , these temporal differences between CD8 and CD4 epitope display held true for all eight EBNA3B epitopes tested ( data not shown ) . The existence of EBNA1 sequence variation between geographically distinct EBV isolates [26] allowed us to generate LCLs using a Chinese virus strain ( CKL ) with epitope mutations that , for the T cell clones used in these experiments , abrogated CD8 recognition and reduced CD4 recognition to a very low base-line . Into these LCLs , we then introduced an epitope-positive EBNA1 allele under dox-regulated control . As shown in Figure 3 , we used both a full length EBNA1 sequence and a sequence ( E1dGA ) from which the internal glycine-alanine repeat ( GAr ) domain had been deleted . Note that this GAr domain reportedly offers the wild-type protein some level of protection from CD8+ T cell recognition through reducing the rate of its translation from mRNA [30] and/or though stabilising the protein from proteasomal digestion [31] . Figures 3A and B show immunoblots of EBNA1 expression induced in the pEBNA1-tet and pE1dGA-tet LCLs following 100 ng/ml dox induction . As with inducible EBNA3B , the two forms of EBNA1 accumulated to reach their steady state levels by 72–96 h post-induction , though E1dGA was detectable slightly earlier than full length EBNA1 ( 6 versus 12 h post-induction ) , and accumulated to slightly higher steady-state levels , a finding consistent with published data [30] , [32] . We examined the kinetics of EBNA1 and E1dGA presentation using clones against two CD8+ ( HPV/B*3501 and IPQ/B*07 ) and two CD4+ ( GLR/DQB1*06 and VYG/DRB1*11 ) T cell epitopes ( see Table S1 ) . Results from one such set of assays are shown in Figures 3C and D , using HPV- and VYG-specific effectors and target LCLs established from a B*3501 , DRB1*11-positive donor . Focusing first on the CD8+ T cell data , we found that both EBNA1 and E1dGA were rapidly recognised by CD8+ T cells and reached their plateau values ( shown by long-term induced cells ) within 48 h . Note that these plateau values were always some 20–30% higher with target cells expressing the E1dGA construct . Given the reported effect of the GAr domain on MHC I processing , we looked in greater detail at early time points in the above experiment , repeating the CD8 assays hourly over the first 12 hr post-induction . As illustrated in Figure S1 for assays conducted with 25 and 100 ng/ml dox inductions , we found that CD8 epitope display from E1dGA was indeed slightly accelerated at early times , typically reaching 35% of its plateau value by 12 hr compared to 25–30% for full length EBNA1 . Turning now to the CD4+ T cell data in Figures 3C and D , antigen presentation by the MHC II pathway was again profoundly delayed . Thus there was no CD4+ T cell recognition of dox-induced target LCLs ( other than very weak base-line recognition of the CKL virus-coded EBNA1 ) until 48 h post-induction , followed by a slow rise that did not reach the long-term plateau value even by 168 h . Both the EBNA1 and E1dGA proteins gave similar results in this respect , although here the plateau level of CD4+ T cell recognition was always slightly higher with cells expressing the full length protein . Experiments conducted on a different pair of pEBNA1-tet and pE1dGA-tet LCLs using T cell clones specific for the IPQ/B*07 and GLR/DQB1*06 epitopes gave the same pattern of results ( data not shown ) . The temporal differences between CD8 and CD4 epitope display therefore held true for all epitopes studied both in EBNA3B and in EBNA1 . However we reasoned that the delayed presentation of CD4 epitopes might simply reflect their processing by an indirect route if , as previously shown for EBNA3A and 3C , the source antigens access the MHC II pathway through antigen release and uptake by neighbouring cells in the LCL culture [20] . We first investigated this for EBNA3B by co-cultivating “antigen donor” cells ( a pEBNA3B-tet LCL lacking relevant MHC restriction alleles but dox-induced to express cognate antigen ) with “antigen-recipient” cells ( an antigen-negative EBNA3B-KO LCL with the relevant MHC alleles ) for 7 days , then used this mixture as a target for EBNA3B-specific CD4+ and CD8+ T cell clones . As shown in Figure S2A , we found that co-culture could indeed sensitise recipient cells to recognition by CD4+ T cell clones specific for the EBNA3B ILR epitope , although not by the corresponding CD8+ IVT clones . However , in parallel experiments where we co-cultured dox-induced pEBNA1-tet “antigen donor” cells with a CKL virus-transformed “antigen recipient” LCL , there was never any recognition of the co-culture by EBNA1-specific CD4+ T cells ( Figure S2B ) . Furthermore a second sensitive method of detecting inter-cellular antigen transfer , where recipient cells are fed with 25x-concentrated culture supernatant from donor LCLs [20] , again never sensitised recipient cells to EBNA1-specific effectors ( Figure S2C ) . This clearly shows that inter-cellular antigen transfer likely contributes to EBNA3B's presentation via the MHC II pathway in LCL cells; however , as others have also observed [17] , endogenously expressed EBNA1 is presented by an intracellular route . Yet , irrespective of these differences , both antigens show delayed presentation via the MHC II pathway following the induction of antigen synthesis . We therefore sought reassurance that this slow presentation did not simply reflect an intrinsic feature of MHC class II maturation and epitope display in our LCL cells . To do so , we used the inducible vector system to express E1dGA fused with an invariant chain ( Ii ) tag that delivers the protein directly into endosomes and the MHC II processing compartment [33] . As shown in Figure 4A , expression of the E1dGA-Ii protein is detectable by immunoblotting 24 , 48 and 72 h after 100 ng/ml dox-induction but at very low levels compared to non-tagged EBNA1 and E1dGA . This reflects on-going degradation of the endosomally-targeted E1dGA-Ii protein , since adding chloroquine , an inhibitor of endosomal proteolysis , 24 h prior to harvest increased the level of protein detectable . Figure 4B shows the corresponding T cell assay data following dox-induction . The Ii-tagged protein was rapidly presented not just to CD8+ T cells , where it was processed as quickly as the non-targeted constructs , but also to CD4+ T cells . In this latter case , recognition appeared within 12 h and became almost maximal by 48 h , much quicker than with the non-tagged proteins . Thus our LCLs can rapidly process and present endogenously expressed antigen , once that antigen gains access to the MHC II presentation pathway . We then examined antigen presentation in long-term 100 ng/ml dox-induced cells after switching off new antigen synthesis by dox-withdrawal . As illustrated in Figure 5A , using a Q-RT-PCR assay for vector-encoded EBNA3B mRNA transcripts , we first showed that >80% of transcripts are lost within 6 h and none are detectable by 24 h . This implies that new antigen synthesis must terminate quite rapidly after dox withdrawal . However , as shown in Figure 5B , the EBNA3B protein is clearly very stable since it remained easily detectable in immunoblots for several days post-withdrawal . Indeed , as the immunoblots were loaded with equal number of cells each time , the falling EBNA3B levels reflect both slow natural turnover of the protein and also dilution from cell doubling ( in cultures with a doubling time of 48–72 hr ) . Aliquots of LCL cells from the same experiment ( HLA B*2702 , DRB3*0201-positive ) were used in parallel as targets for EBNA3B-specific T cells . As shown in Figure 5C , target cell recognition by an RRA epitope-specific CD8+ T cell clone fell progressively after dox withdrawal , down to half of the original level by 48 h , to <10% by 96 h and approaching zero thereafter . By contrast , recognition by a CD4+ T cell clone against the QAP epitope fell much more slowly , being still >50% of the original level after 96 h and >20% even after 192 h . Indeed the rate of fall in CD4 epitope display closely paralleled the level of EBNA3B protein detectable in these target cells by immunoblotting ( cf . Figures 5B and 5C ) . Such experiments were conducted on all three pEBNA3B-tet LCL backgrounds , whether first induced at 25 ng/ml or 100 ng/ml dox , and included clones against five CD8 epitopes and three CD4 epitopes . In each case CD8+ T cell recognition had fallen to <10% of its original value by 96 h after dox withdrawal , whereas CD4+ T cell recognition was still at 35–50% of its original value at the much later time of 168 h ( data not shown ) . Results from a corresponding experiment involving pEBNA1-tet and pE1dGA-tet LCLs are shown in Figure 6 . Q-RT-PCR assays using primer/probe combinations specific for vector-encoded EBNA1 and E1dGA mRNAs showed mRNA levels fell rapidly after dox-withdrawal and were undetectable beyond 12 h ( Figure 6A ) . Again , therefore , new antigen synthesis must rapidly terminate following dox withdrawal yet , as shown by the immunoblots in Figures 6B and 6D , both the EBNA1 and E1dGA proteins are relatively stable , levels per cell falling slowly over time and being still detectable at 168 h . When these same dox-withdrawn cells ( HLA B*3501 , DRB1*11-positive ) were used as targets in T cell assays , recognition by HPV-specific CD8+ T cells fell to <50% of the original level by 48 h and was undetectable by 120 h , whereas recognition by a VYG-specific CD4+ T cells fell much more slowly , being still 30–40% of the original value as late as 168 h . Again , parallel experiments using a different LCL background and T cell clones against the other CD8 and CD4 epitopes in EBNA1 produced a very similar pattern of results . While Figures 5 and 6 showed that CD8 and CD4 epitope display fell at different rates after switching off new antigen synthesis , in both cases target cells remained susceptible to T cell recognition for some time . We therefore asked how the observed rates of fall compared to the half-lives of pre-existing MHC I-peptide and MHC II-peptide complexes on the LCL surface . Thus pEBNA3B-tet and pEBNA1-tet LCLs of the appropriate MHC type maintained in the absence of dox were briefly exposed to a non- saturating dose of epitope peptide , washed well ( time 0 h ) and the subsequent fall in epitope display tracked over time by T cell assay . For comparison , all experiments included long-term-induced cultures of the same LCLs , from which dox was either withdrawn at time 0 h or maintained throughout . Figure 7 shows representative data obtained for pairs of epitopes from EBNA3B and from EBNA1 . Both CD8 epitopes had half-lives on the LCL surface of 36–48 h; indeed the rate with which exogenously loaded CD8 peptides disappeared from the surface was only slightly faster than the rate at which CD8 epitope display fell following cessation of new antigen synthesis . However , both CD4 epitopes also had half-lives in the same range , the levels of display on peptide-pulsed cells therefore falling much quicker than seen on pEBNA3B-tet and pEBNA1-tet LCL cells after cessation of antigen synthesis . A similar pattern of results was observed for all CD8+ and CD4+ T cell epitopes tested ( see for example Figure S3 ) . Such results strongly suggest that , after the cessation of antigen synthesis , new CD4 epitope complexes continue to reach the cell surface whereas the supply of new CD8 epitope complexes is rapidly curtailed . To test this further , we used a protocol ( briefly exposing cells to citrate/phosphate buffer at pH 3 . 1 ) that efficiently strips pre-existing EBV epitope/MHC I and/MHC II complexes from the LCL surface without affecting cell viability . Having switched off antigen synthesis in pEBNA3B-tet and pEBNA1-tet LCLs by dox withdrawal , we followed the recovery of epitope peptide display by T cell recognition , stripping pre-existing epitopes off the cell surface either at the time of dox withdrawal ( time 0 h ) or 48 h later . The results from such assays are illustrated in Figure 8 , again comparing CD8/CD4 epitope pairs from EBNA3B and from EBNA1 . In each case , new epitope supply after stripping at time 0 h ( blue line ) or 48 h ( red line ) is shown against the level of surface epitope display seen on the same target cells that had been similarly dox-depleted at time 0 h but not stripped ( black line ) . The CD8 epitopes showed significant recovery of cell surface display 24 h after stripping at time 0 h but then levels fell away rapidly , down to the same low values remaining on dox-depleted , non-stripped cells . When stripping was delayed until 48 h after dox-withdrawal , there was only a small recovery of CD8 epitope display , recapitulating the low residual values on non-stripped cells . By contrast , the CD4 epitopes showed a substantial recovery whether the cells were stripped at 0 h or 48 h following dox-withdrawal . Furthermore the recovery was sustained for up to 192 h , with stripped cells regaining the same persistent levels of CD4 epitope display as shown by non-stripped cells . Figure S4 shows the results of a similar experiment involving different target LCLs , here initially induced at 25 ng/ml dox , and effectors against different epitopes . This emphasises the point that consistent results were obtained for all CD8/CD4 epitope pairs tested , whether antigen was initially expressed at physiological or supra-physiological levels .
Here we address a generic question regarding pathogens , particularly viruses , that naturally infect target cells in which both the MHC I and MHC II pathways of antigen presentation are active . Antigens endogenously expressed within such an infected cell could potentially be presented by both pathways , rendering the cell visible to CD4+ as well as CD8+ T cells . However , the relative timing of those events and their degrees of dependence upon new antigen synthesis have never been rigorously examined in parallel . Our experimental system , based on EBV-infected B cell lines and the regulatable expression of EBV antigens , allows one to study these processes in a physiologically relevant cell context , select appropriate levels of antigen expression and track the presentation of CD8 and CD4 epitopes from the same source antigen after inducing or suppressing antigen synthesis . We studied five CD8 and three newly-defined CD4 epitopes from EBNA3B and two CD8 and two CD4 epitopes from EBNA1 , in each case probing epitope display with at least two independent clones per epitope . To cover the wide range of MHC restricting alleles involved , assays were conducted on three different pEBNA3B-tet LCLs and two different pairs of pEBNA1-tet and pE1dGA-tet LCLs . The contrasting patterns of CD8 versus CD4 epitope display were remarkably consistent across the whole range of epitopes and antigens studied , and were reproducible whether the antigen was being expressed at physiologic ( LCL-like ) or supra-physiologic levels . We infer that these differences are not chance consequences of particular epitope selection but reflect fundamental differences in the way that endogenously expressed viral antigens are handled by the MHC I and MHC II presentation pathways in human B cells . At the same time , we would emphasise that both EBNA3B and EBNA1 are native nuclear proteins; there could possibly be differences in detail were one to study the processing of viral antigens normally resident in the cytoplasm or marked for export , but we would nevertheless expect the basic pattern of results to remain the same . With antigen induction , we found that EBNA3B and both forms of EBNA1 were rapidly recognised by CD8+ T cells . Recognition was first apparent soon after dox addition and rose to almost maximal levels within 24 h , well before steady-state levels of these proteins , as detected by immunoblotting , were reached . The results with EBNA1 were particularly interesting given the history of work on this protein as a target for CD8+ T cells . Thus early studies found that the GAr domain was able to protect EBNA1 from presentation via the MHC I pathway [31] and that this was associated with resistance to proteasomal degradation [31] , [34] . However , more recent results have shown that this protection from CD8+ T cell recognition is only partial [35] , [36] , [37] , [38] and may reflect a GAr-mediated reduction of the rate of protein translation rather than of sensitivity to the proteasome per se [30] , [32] , [39] , [40] . Importantly , many of these studies involved chimaeric antigen constructs , often with indicator epitope insertions , tested in in vitro translation or transient transfection assays , leaving the effects of the GAr domain in its physiologic setting open to question . In the present work we found that , after inducing antigen synthesis , E1dGA was presented to CD8+ T cells slightly quicker than the wild-type protein , though the magnitude of the effect was not as great as noted in other less physiologic experimental settings . We believe that our system is robust in this regard since we also found that CD8+ T cell recognition of cells induced to express E1dGA long-term was consistently 20–30% greater than seen with cells induced to express EBNA1 . This exactly mirrors levels of EBNA1 epitope display seen earlier in LCLs transformed with EBV expressing a GAr-deleted EBNA1 protein versus LCLs transformed with wild-type virus [35] . Overall the results of the antigen induction experiments were consistent with MHC I presentation of newly synthesised polypeptides . However , in the same experiments , the MHC II-restricted presentation of EBNA3B and EBNA1 was grossly delayed; CD4 epitope display only became detectable after 36–48 h and took some 7 days to reach the long-term steady state level . This delay is not an intrinsic feature of MHC II processing in LCL cells since an invariant chain-targeted E1dGA protein expressed in the same dox-inducible system was detected by CD4+ T cells within 12 h and maximum recognition was reached within 48 h . This reinforces a large amount of earlier evidence testifying to the efficiency of MHC II antigen processing in LCL cells [41] . Our findings therefore imply that endogenously expressed antigens such as EBNA3B and EBNA1 are delivered very slowly into the MHC II processing pathway , even though they may access that pathway by different routes . Thus co-cultivation experiments showed that EBNA3B ( like EBNA3A and 3C , [20] ) is processed , at least in part , via the inter-cellular transfer of antigen between LCL cells . The precise form of antigen being transferred in LCL cultures is not known , except that it clearly requires active processing and , by analogy with our earlier work using donor cells transformed with a replication-defective EBV strain [20] , does not derive from cells dying as a result of lytic virus replication . By contrast , the same experimental approaches never detected any evidence for inter-cellular transfer of EBNA1 . Thus the CD4 epitopes recognised by our EBNA1-specific T cell clones must derive from antigen processed by an intracellular route . In that regard , others have also observed that endogenously expressed EBNA1 is processed intracellularly in LCL cells , and have suggested the involvement of autophagy in that process [17] . Dox-withdrawal from pre-induced LCLs allowed us to ask whether , in the absence of new antigen synthesis , the pre-formed intracellular pool of mature protein can feed the MHC I and MHC II pathways . We first verified that gene transcription from the dox-inducible promoter terminated rapidly after dox-withdrawal , with EBNA3B and EBNA1 transcript levels falling by >80% within 6 h and becoming undetectable by 12-24 h . New antigen synthesis must therefore be curtailed at least at the same rate yet , as is clear from the immunoblots , the pre-formed EBNA3B , EBNA1 and E1dGA proteins remain detectable for days thereafter . In this regard the natural turnover of EBNA3B has not been investigated previously , while ours is the first attempt to compare turnover of the EBNA1 and E1dGA proteins having switched off their synthesis specifically , rather than non-specifically with general protein synthesis inhibitors . Previous studies of the latter kind , where EBNA1 is first expressed by transient transfection or from recombinant viral vectors , all indicate that the wild-type protein has a long half-life , but differ in the degree to which this is shortened by GAr deletion [31] , [32] , [36] , [42] . Our finding , that in the natural setting of the LCL cell both EBNA1 and E1dGA are stable proteins , accords with the most recent findings from transiently transfected cells with protein synthesis inhibitors [42] . For our present purpose , however , the essential point is that both EBNA3B and the two forms of EBNA1 are sufficiently stable that a large pool of mature protein persists in the cells for several days after the cessation of new antigen synthesis , providing a source of antigen that is potentially available to both MHC I and MHC II presentation pathways . It is therefore significant that , upon dox-withdrawal , T cell assays showed a marked fall in cell surface display of all seven CD8 epitopes tested , typically to <50% of the initial level by 48 h and to <10% by 96 h . Indeed the rate of fall was in each case close to that seen when the corresponding epitope-negative LCL cells were loaded with epitope peptide at non-saturating levels and tracked over time to follow the natural half-life of the MHC I-peptide complex on the cell surface . These half-life measurements accord with earlier work , for example the RRA/B*2702 epitope from EBNA3B was estimated to have a half-life of 40 h in the present T cell assays and of 37 h in earlier antibody-based assays [43] . While rates of fall were similar on dox-depleted and peptide-pulsed cells , there was often a slight delay in the timing of that fall on dox-depleted cells . At least part of this lag must reflect the fact that , for a short time after dox-withdrawal , new MHC-peptide complexes either already in the export pathway or generated from residual mRNA translation will be delivered to the cell surface . Overall , the results strongly suggest that continued CD8 epitope display depends upon continued antigen synthesis . By contrast , T cell recognition of CD4 epitopes consistently fell much more slowly after dox-withdrawal , typically being still >50% of the initial level at 96 h and still easily detectable as late as 192 h . Recognition persisted despite the fact that in peptide pulsing experiments the relevant MHC II/CD4 epitope complexes have half-lives similar to their MHC I/CD8 epitope counterparts , strongly implying that the MHC II presentation pathway was being fed from the mature protein pool . These conclusions were further supported by experiments in which cells were stripped of cell surface peptides after dox-withdrawal , and then assayed for the recovery of epitope display over time . Interestingly , cells stripped immediately after dox withdrawal showed a significant recovery of detection by CD8+ T cells 24 h later; however this effect , which could be quite marked for some epitopes , was transient with recognition falling away at later times . We attribute this transient recovery to the continued supply of newly-formed complexes to the cell surface occurring immediately after dox-withdrawal ( as above ) and possibly also to the reappearance of pre-existing mature complexes that were recycling from the surface at the time of stripping [44] . Importantly , cells stripped 48 h after dox removal , by which time surface epitope display was declining rapidly , showed minimal recovery of CD8 recognition . This strongly suggests that the mature protein pool , which is still substantial in cells 48 h after dox-withdrawal , makes little if any contribution to the MHC I presentation pathway . By contrast , CD4 epitope display was extensive and prolonged , whether cells were stripped immediately after dox-withdrawal or 48 h later . Such sustained presentation of CD4 epitopes by cells in which de novo synthesis of EBNA3B , EBNA1 and E1dGA was terminated 48 h earlier must reflect processing of antigen derived from the mature protein pool . In summary , we find that in virally-infected human B cells newly-synthesised viral polypeptides , by inference rapidly degraded translation products , are the dominant source of antigen feeding the MHC class I pathway . This does not discount the possibility that the mature protein pool may , in other circumstances or in other cell types , contribute to such a role . Indeed , prompted by a report that irradiation could increase MHC I processing activity in cycloheximide-treated cells [45] , we irradiated pEBNA3B-tet and pEBNA1-tet LCLs several days after dox-withdrawal and showed a small , transient recovery of CD8 epitope display that , in the absence of antigen synthesis , must have come from mature protein ( L . K . Mackay , unpublished observations ) . However we find no evidence of any major contribution to the MHC I pathway from this source in a naturally proliferating LCL cell . By contrast , in these same cells endogenous antigen presentation via the MHC II pathway is dependent upon the mature protein pool and shows no immediate connection with the presence or absence of de novo translation products . These fundamental differences have important implications for virus-specific CD8+ and CD4+ T cells as direct effectors against infections of MHC I/II-positive target cells . In such circumstances , only CD8+ T cells have the capacity to recognise newly infected cells as soon as de novo antigen synthesis begins; CD4+ T cell recognition will be delayed until the intracellular antigen pool has increased sufficiently to feed the MHC II presentation pathway . Interestingly however , our results imply that for viruses establishing latent infections in MHC I/II-positive cells where viral gene expression is extinguished but where viral proteins persist , a situation that could for example pertain to gamma-herpesviruses and their genome maintenance proteins , the latently-infected cell reservoir may remain visible to CD4+ T cells .
All experiments were approved by the South Birmingham Local Research Ethics Committee ( 07/Q2702/24 ) . All patients provided written informed consent for the collection of blood samples and subsequent analysis . LCLs were established using the reference EBV strain B95 . 8 , a B95 . 8-based recombinant lacking the EBNA3B gene ( EBNA3B-KO ) [23] , or the Chinese CKL strain ( called NPC 15 , [46] ) with a variant EBNA1 sequence . All lines were maintained in RPMI 1640 medium supplemented with 2 mM glutamine and 10% fetal calf serum ( standard medium ) . A derivative of the dox-dependent expression vector pRTS-1 [27] and the EBNA3B , EBNA1 and E1dGA constructs were kindly provided by Dr J Mautner , Munich; in cases where EBNA1 was expressed under dox control , a derivative of pRTS-1 lacking constitutively expressed EBNA1 was used . Ii-tagged E1dGA and FLAG-tagged EBNA1 and E1dGA constructs were constructed by PCR , verified by DNA sequencing , then introduced into the vector by standard DNA cloning procedures . To introduce these into LCLs , DNA ( 15 µg ) was transfected into 107 cells by electroporation in 300 µl Optimem ( Invitrogen ) at 230 V and 960 µF using a Biorad electroporation apparatus . Immediately after electroporation , cells were resuspended in RPMI 10% FCS and were incubated at 37°C and 5% CO2 . After 24 h in culture , cells were then stained with rat CD2-specific antibody OX34 and were positively selected by magnetic cell sorting with anti-mouse IgG2a/b Microbeads and LS columns ( Miltenyi Biotech ) according to the manufacturer's guidelines . Cells were then expanded and maintained in culture in the absence of dox , before testing for dox- inducibility of antigen expression . Total RNA was extracted from 5×106 cells using a Nucleospin RNA extraction kit ( Macherery Nagel ) according to the manufacturer's instructions . 400 ng RNA was reverse transcribed into cDNA using a pool of primers specific for EBNA3B , EBNA1/E1dGA and ( as an internal control ) cellular GAPDH transcripts . In subsequent quantitative PCR ( Q-PCR ) assays , primer/probe combinations were used to amplify ( i ) the 3′ end of the major EBNA3B exon , or ( ii ) the unique 5′ end of EBNA1/E1dGA transcripts initiated from the dox-regulatable promoter in plasmid pEBNA-tet . After normalising to GAPDH expression , levels of EBNA3B or EBNA1/E1dGA transcription in test cells are expressed relative to that of a fully induced cell line . Cells were sonicated in UTB buffer ( 8 M urea , 150 mM β-mercaptoethanol , 50 mM Tris/HCl pH 7 . 5 ) and cellular debris removed by centrifugation . Protein concentration was determined by using the BioRad Bradford Protein determination reagent . Solubilized proteins were separated by 8% SDS-PAGE and transferred to nitrocellulose membranes ( Thermo Scientific Pierce ) . Cellular and viral proteins were detected by incubating the membranes with specific Abs followed by HRP-conjugated secondary Abs ( Sigma ) . Bound HRP was visualized using the ECL-plus detection kit ( Amersham Biosciences ) . Antibodies used include: anti-EBNA3B ( ExAlpha ) , anti-EBNA1 ( IH4 , [47] ) , and anti-actin ( Sigma ) . CD4 epitope peptides within EBNA3B were identified by screening immune donor lymphocytes in IFNγ Elispot assays on peptide panels ( 20-mers overlapping by 15 ) covering the primary sequence of B95 . 8 strain EBNA3B . All peptides were synthesized using 9-fluorenylmethoxycarbonyl chemistry ( Alta Bioscience; University of Birmingham , Birmingham , U . K . ) , dissolved in DMSO , and concentrations were determined by biuret assay . CD4+ and CD8+ T cell clones specific for these and for other defined epitopes within EBNA 1 or EBNA3B were generated as described [29] . All epitope sequences are shown , with their MHC restricting alleles , in Table S1 . Immediately before all T cell assays , target LCL cells were fixed in 1% paraformadehyde for 10 min followed by quenching with 0 . 2 M glycine for 10 min , and then washed with PBS before resuspension in standard medium . Assays therefore measured the level of epitope display at a defined time point , with no further changes occurring during the 18 h assay period itself . Unless otherwise stated , fixed target cells were seeded at 105 cells per triplicate assay well , to which 2000 T cells were added; after 18 h incubation , supernatant medium was harvested and assayed for IFNγ release by ELISA ( Endogen ) . Assays routinely included the following control targets: the wild-type B95 . 8 virus-transformed LCL from the same donor as the pEBNA-tet-transfected LCLs , the relevant pEBNA-tet-transfected LCL both without dox induction and long-term dox-induced , and the pEBNA-tet-transfected LCL without dox-induction but exogenously loaded with 10−7 M concentration of the relevant epitope peptide . In all assays , at least two different T cell clones were tested for each epitope specificity . In assays measuring the half-life of peptide/MHC complexes at the cell surface , LCLs with relevant HLA types but transformed with EBNA3B-KO or CKL ( variant EBNA1 ) virus strains were exposed for 1 h to epitope peptide at concentrations mediating half-maximal recognition , then washed several times and either fixed immediately for T cell assay , or cultured in standard medium then harvested and fixed for assay at later times . For assays measuring the continued supply of complexes to the surface from intracellular sources , we used an acid-stripping protocol that preliminary work confirmed would completely remove both MHC I- and MHC II-bound epitope peptides without affecting cell viability ( [35] and L . Mackay , unpublished observations ) . Cells were washed with PBS and pellets were gently resuspended in citrate/phosphate buffer ( 0 . 131 M citric acid , 0 . 066 M Na2HPO4 ) , pH 3 . 1 , for 20 min on ice before neutralization by addition of excess standard medium . Stripped target cells were then washed several times and an aliquot of cells fixed immediately for T cell assay , while the remaining cells were re-cultured in standard medium , then harvested and fixed for assay at later times .
|
Many viruses infect cells in which both the MHC I and MHC II pathways of antigen presentation are active , and so viral proteins expressed in those cells may be presented as MHC I-peptide complexes to CD8+ T cells and as MHC II-peptide complexes to CD4+ T cells . Here we study these processes in a model system involving Epstein-Barr virus-infected human B lymphocytes ( MHC I/II-positive ) where viral antigen expression can be induced or suppressed at will , and antigen presentation tracked with specific CD8+ and CD4+ T cell clones . In this system , we find that the MHC I pathway is entirely fed by newly-synthesised polypeptides , whereas the MHC II pathway depends upon antigen supplied from the mature protein pool . Hence , while only CD8+ T cells can rapidly recognise new infections , only CD4+ T cells will recognise latent infections in which viral gene expression is extinguished yet a pool of viral antigens remains .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"virology/persistence",
"and",
"latency",
"immunology/antigen",
"processing",
"and",
"recognition",
"virology/viral",
"replication",
"and",
"gene",
"regulation",
"virology/vaccines",
"immunology/immunomodulation",
"immunology/immune",
"response",
"immunology/immunity",
"to",
"infections",
"virology/new",
"therapies,",
"including",
"antivirals",
"and",
"immunotherapy",
"virology/viruses",
"and",
"cancer",
"virology/immune",
"evasion",
"virology/host",
"invasion",
"and",
"cell",
"entry",
"oncology/hematological",
"malignancies",
"virology/host",
"antiviral",
"responses"
] |
2009
|
T Cell Detection of a B-Cell Tropic Virus Infection: Newly-Synthesised versus Mature Viral Proteins as Antigen Sources for CD4 and CD8 Epitope Display
|
A family of secreted cathepsin L proteases with differential activities is essential for host colonization and survival in the parasitic flatworm Fasciola hepatica . While the blood feeding adult secretes predominantly FheCL1 , an enzyme with a strong preference for Leu at the S2 pocket of the active site , the infective stage produces FheCL3 , a unique enzyme with collagenolytic activity that favours Pro at P2 . Using a novel unbiased multiplex substrate profiling and mass spectrometry methodology ( MSP-MS ) , we compared the preferences of FheCL1 and FheCL3 along the complete active site cleft and confirm that while the S2 imposes the greatest influence on substrate selectivity , preferences can be indicated on other active site subsites . Notably , we discovered that the activity of FheCL1 and FheCL3 enzymes is very different , sharing only 50% of the cleavage sites , supporting the idea of functional specialization . We generated variants of FheCL1 and FheCL3 with S2 and S3 residues by mutagenesis and evaluated their substrate specificity using positional scanning synthetic combinatorial libraries ( PS-SCL ) . Besides the rare P2 Pro preference , FheCL3 showed a distinctive specificity at the S3 pocket , accommodating preferentially the small Gly residue . Both P2 Pro and P3 Gly preferences were strongly reduced when Trp67 of FheCL3 was replaced by Leu , rendering the enzyme incapable of digesting collagen . In contrast , the inverse Leu67Trp substitution in FheCL1 only slightly reduced its Leu preference and improved Pro acceptance in P2 , but greatly increased accommodation of Gly at S3 . These data reveal the significance of S2 and S3 interactions in substrate binding emphasizing the role for residue 67 in modulating both sites , providing a plausible explanation for the FheCL3 collagenolytic activity essential to host invasion . The unique specificity of FheCL3 could be exploited in the design of specific inhibitors selectively directed to specific infective stage parasite proteinases .
The common liver fluke F . hepatica , together with F . gigantica , are the causative agents of fascioliasis , a zoonosis causing huge global losses in the agricultural section by infecting more than 700 million ruminants worldwide . The disease is also recognized by the WHO as an important emerging neglected disease of humans , particularly in areas of South America , Asia , Iran and Egypt [1] . Infection with this parasite is acquired by the ingestion of plants contaminated with metacercariae , a resistant cystic form that emerges as a newly excysted juvenile ( NEJ ) in the duodenum , and after traversing the gut wall migrates to the liver . The parasites spend 8–12 weeks feeding on , and severely damaging , the liver parenchyma before they move into the bile ducts and become obligate blood-feeders by sucking blood through punctures in the duct walls . As in other parasites the invasion and establishment is mediated by a delicate crosstalk between molecules generated by the parasite and the host , with proteolytic enzymes being major players in this interaction [2] . Tissue migration and feeding is facilitated by the abundant secretion of proteolytic enzymes , most particularly cathepsin L cysteine proteases [3] , [4] . F . hepatica possesses an expanded multigene family of cathepsin L-like proteases that includes at least 5 different Clan CA ( papain-like ) members that are developmentally regulated and play pivotal roles in parasite survival by facilitating migration , immune evasion and feeding [5] , [6] . Transcriptomic and proteomic studies have demonstrated that the infective NEJ express and secrete cathepsin L3 ( FheCL3 ) indicating that this is critical to enabling the parasite penetrate the intestinal wall [7] , [8] , [9] , [10] . By contrast , the blood-feeding adult expresses predominantly cathepsinL1 ( FheCL1 ) , to a lesser extent , cathepsin L2 ( FheCL2 ) and to a relatively minor extent FheCL5 . FheCL1 can be involved in parasite feeding , since in vitro experiments showed it can digest hemoglobin; both FheCL1 and FheCL2 have been implicated in immune evasion based in their in vitro ability to cleave native immunoglobulins [11] . Correlating with the macromolecular substrates the parasite encounters at these different locations , the cathepsin L members exhibit distinct substrate specificities [4] , [11] . For papain-like proteases , the evidence points to the S2 subsite as being most critical to defining substrate selectivity [12] . We have shown that the juvenile FheCL3 is unusual in having a particular preference for Pro residues in the P2 position of peptide substrates . By stark contrast , FheCL1 has a marked preference for aliphatic and aromatic residues in the P2 substrate position and does not readily accept Pro . FheCL2 , on the other hand , exhibits an substrate preference in between these two enzymes by preferring P2 aliphatic and aromatic residues but also accepting Pro , although much less efficiently than FheCL3 . Most interestingly , we have previously demonstrated that the preference for P2 Pro confers FheCL3 and FheCL2 with the rare ability to cleave native collagen [13] , [14] . Only two other cysteine proteases , mammalian cathepsin K , which is involved in bone resorption by osteoclasts [15] , and the ginger rhizome cysteine proteases ( CP-II or zingipain , GP2 and GP3 ) also exhibit this high affinity for Pro in P2 and collagenolytic activity [16] , [17] . Comparison of crystallographic structures of several Clan CA cysteine proteases allowed the identification of residues that make up the active site cleft with the selective S2 pocket being delimitated by residues 67 , 133 , 157 , 158 and 205 ( papain numbering ) [18] , [19] , [20] , [21] , [22] , [23] , [24] . While variations occur in several of these positions within the F . hepatica cathepsin L family the residue at position 67 has been primarily implicated in P2 Pro accommodation by stabilizing interactions with the planar ring of Pro in the peptide substrate [20] , [25] . In FheCL3 and zingipain this position is occupied by the large aromatic residue Trp while in FheCL2 and cathepsin K a Tyr is present . Structural comparisons and molecular dynamic simulations performed by us suggested that the substrate selectivity observed in FheCL3 might be due to steric restrictions imposed by the bulky aromatic residues not only at the S2 subsite but also within the S3 pocket [13] , [14] . The remarkable convergence between FheCL3 and zingipain is not only restricted to Trp67 but also the close-by position 61 at the bottom of the S3 pocket is occupied by a large His residue . This suggested to us that together these two active site moieties could influence the capacity of the enzymes to best accommodate Pro over other aliphatic residues , and hence account for their collagenolytic activity . To get a clear picture of the substrate specificity of the major proteases of F . hepatica , we used a recently developed method involving multiplex substrate profiling and mass spectrometry ( MSP-MS ) , that provides for unbiased subsite profiling of proteases across the entire active site [26] . In addition , the P1–P4 subsite specificities were determined by Positional Scanning- Synthetic Combinatorial Libraries of fluorogenic tetrapeptides ( PS-SCL ) , a well-established technology to study protease substrate specificity [20] , [27] , [28] , [29] . To test the relevance of active site positions 61 and 67 in selectivity we prepared recombinant variants of FheCL3 with the specific alterations in the S2 and S3 subsites , mimicking those present in FheCL1 , and the reciprocal variants of FheCL1 in an attempt to confer this protease with collagenolytic activity . All the approaches highlight the unusual and marked preference of FheCL3 for P2 Pro , and additionally reveal that the P3 pocket has a less marked but distinctive preference for Gly . The mutational analysis emphasizes the dual role of residue 67 in modulating interactions with both P2 and P3 substrate residues and its crucial importance in juvenile FheCL3 specificity and activity . Our findings provide structural insights into the molecular determinants of active site preferences of two proteases that are vital for parasite development , which might in turn prove useful in the design of strategies to control parasite infection .
Six FheCL3 and FheCL1 variants bearing substitutions at the S2 and S3 active site pockets were constructed by site-specific mutagenesis using the QuikChange Site-Directed Mutagenesis Kit ( Stratagene ) as indicated in Table S1 . Briefly , different pairs of complementary oligonucleotides containing the base pair substitutions to be introduced in the cathepsin gene sequences were generated and used in an outside PCR reaction employing as templates clones of FheCL1 or FheCL3 in the X4-Mfα-ScPas3 expression plasmid ( kindly provided by Dr . R . J . S . Baerends and Dr . J . A . K . W . Kiel , Molecular Cell Biology Lab , Groningen Biomolecular Sciences and Biotechnology Institute , University of Groningen , The Netherlands ) . Double variants were obtained by using plasmids bearing the single mutations as templates . The amplified modified plasmids were propagated in bacteria , sequenced to confirm the presence of the desired mutations , and then electroporated in the Hansenula polymorpha yeast strain for production as previously described [30] . FheCL1 and FheCL3 recombinant proenzymes were produced in the yeast Hansenula polymorpha as previously described [13] , [14] . Briefly , yeast transformants were cultured in 500 ml BMGY broth at 37°C to an OD600 of 2–6 , harvested by centrifugation at 2000 g for 10 min and induced by resuspending in 50 ml of buffered minimal media ( 0 . 67% yeast nitrogen base; 0 . 1M phosphate buffer pH 6 . 0;1% methanol ) for 36 hs at 30°C . Recombinant propeptidases were secreted to the culture media , and recovered by 20–30 fold concentration of culture supernatants by ultrafiltration with a 10 kDa cut-off membrane . The proenzymes were autocatalytically activated to the mature form by incubation for 2 h at 37°C in 0 . 1 M sodium citrate buffer ( pH 5 . 0 ) with 2 mM DTT and2 . 5 mM EDTA , dialyzed against PBS pH 7 . 3 and stored at −20°C . The protein concentration was assessed by the BCA method [31] . The proportion of functionally active recombinant enzyme was determined by titration against E-64c . The enzymes variants were obtained with the same protocol used for production of FheCL1 and FheCL3 . The enzymatic activity of FheCL1 and FheCL3 were compared by MSP-MS , a procedure designed for unbiased profiling of protease activity [26] . A highly diversified peptide library consisting of 124 synthetic tetradecapeptides containing all possible amino acid pairs and near neighbor pairs , was used to test enzymatic activity . All peptides had unmodified termini and consist of natural amino acids except Met that was substituted by norleucine and Cys omitted because of potential disulfide bond formation . The library was distributed into three pools consisting of 52 , 52 and 20 peptides and diluted to 1 µM in 25 mM sodium phosphate , pH 6 . 0 , 1 mM DTT , 1 mM EDTA . An equal volume of FheCL1 or FheCL3 in the same buffer was added to the peptide pools such that the final concentration of each enzyme was 10 nM . An enzyme-free assay was set up as a control . Assays were incubated at room temperature and aliquots were removed after 5 , 15 , 60 , 240 and 1200 minutes . All reactions were acid quenched to pH 3 . 0 or less with formic acid ( 4% final ) , evaporated to dryness and reconstituted to the original volume in 0 . 1% formic acid . Ten µl of each time point were injected onto a 150×0 . 3 mm Magic C18AQ column ( Michrom Bioresources ) connected to a Thermo Finnigan LTQ ion trap mass spectrometer equipped with a standard electrospray ionization source . Peak lists were generated from the raw files using PAVA software ( UCSF ) and searched against a database consisting of all 124 peptides using Protein Prospector . Newly formed cathepsinL1 or L3 cleavage products were identified by comparison with control assays . The substrate specificities of FheCL1 , FheCL3 and all the variants were determined using a PS-SCL [26] . Assays were performed in 0 . 1 M sodium phosphate buffer pH 6 . 0 , 1 mM DTT , 1 mM EDTA , 0 , 01% PEG-6000 and 0 . 5% Me2SO ( from the substrates ) at 25°C . Aliquots of 12 . 5 nmol in 0 . 5 µl from each of the 20 sub-libraries of the P1 , P2 , P3 and P4 libraries were added to the wells of a 96-well Microfluor-1 flat-bottom plates . The final concentration of each compound of the 8 , 000 compounds per well was 15 . 62 nM in a 100 µl final reaction volume . The assays were performed in triplicate , the reaction was initiated by addition of the enzyme diluted in the above buffer and monitored with a SpectraMax Gemini fluorescence spectrometer ( Molecular Devices ) with excitation at 380 nm , emission at 460 nm and cutoff at 435 nm . Kinetic parameters were determined in a reaction buffer containing 0 . 1 M sodium phosphate buffer , pH 6 . 0 , 1 mM DTT and 1 mM EDTA at 25°C; typically final enzyme concentrations were in the 10−9M range , and the substrate was added after 10 min . of incubation of enzyme in reaction buffer . Enzyme concentration was determined by active-site titration with E-64c . Enzyme activity was monitored by the hydrolysis of 7-amino-4-methyl coumarin ( AMC ) from the synthetic peptide substrates Z-VLK-AMC and Tos-GPR-AMC . Reaction rates with different substrate concentrations ( 5–100 µM ) were measured in duplicate as the slope of the progress curves obtained by continuous recording in a FluoStar spectrofluorimeter at 345 excitation and 440 emission wavelengths , using an AMC standard curve for product concentration calculation . Kinetic constants , kcat and KM , were estimated by non-linear regression analysis of the Michaelis–Menten plot using the OriginPro 6 . 1 software . Protein digestion was analyzed by incubating 10 µg of type I collagen from rat tail ( Sigma ) with 5 µM enzyme in PBS pH 7 . 3 , 1 mM DTT and 1 mM EDTA for different times at 28°C . Digestion reactions were stopped by adding 10 µM of E64c to the reaction mixture . Fragments were separated by SDS-PAGE gels ( 8% acrylamide ) under reducing conditions and stained with Coomassie Brilliant BlueR-250 . Homology models of FheCL3 were generated with SwissModel [32] using as principal template the crystal structure of FheCL1 ( 206× ) . Template and models were superimposed for visualization with Swiss PDBViewer version 4 . 1 . ( http://www . expasy . org/spdbv/ ) [33] Active site residues were identified based on the literature and confirmed by structural alignment with human cathepsin L ( 1MHW ) , human cathepsin K ( 1ATK ) and papain ( 5PAD ) . The FheCL3 rotamers and the W67L mutant were generated with the mutate function in the PDBViewer , and selected based on rotamer score and visual inspection .
MSP-MS is a novel method designed to profile protease activity , based on the cleavage of a library of 124 tetradecapetides , providing theoretically unbiased information on preferences at both sides of the cleavage point [26] . The extended nature of the tetradecapeptides allow a much more natural interaction across the protease active site providing a detailed picture of the contribution of the S and S' subsites that accommodate the substrate . The characteristics of the S' sites are generally poorly known mainly because most substrates used for enzymatic profiling place a fluorophore or chromophore in the P1′ position , a moiety very unlike any amino acid that the enzyme can normally accept in that pocket . FheCL1 or FheCL3 were added separately to the library and all the cleaved peptides were identified at time intervals by mass spectrometry . While both enzymes cleaved at more than 170 sites after one hour incubation , FheCL1 had produced approximately 75% within five minutes of the reaction and >95% by 15 minutes . Notably , compared to FhCL1 , FheCL3 produced relatively fewer cleavages at early time-points , while minor cleavages still occur for up to 20 hours of reaction , indicating differences in the ability to accommodate substrates ( Figure 1 A and Figure S1 ) . Significantly , only approximately half of the cleavage sites identified at any time were cleaved by both enzymes , leaving many that were exclusive for either FhCL1 or FhCL3 ( Figure 1 B , C ) . A good example of this differential cleavage is offered by Peptide#38 where FheCL1 cleaves once between T∧F ( EAWMT∧FIVPPRSAG ) but FheCL3 cleaves twice between W∧M and R∧S ( EAW∧MTFIVPPR∧SAG ) and never cleaves between T∧F even after 1200 minutes incubation ( data not shown ) . The positional analysis indicate that the substrate specificity in both FheCL1 and FheCL3 is dominated by the amino acid at P2 consistent to what is known about clan CA cysteine proteases [12] ( Figure 2 ) . The substrate signature at this position showed that besides aliphatic residues that can be accommodated by both enzymes , FheCL1 can readily accept Phe at P2 but has very low tolerance for Pro , while FheCL3 is the opposite ( Figure 1B–C ) . In fact the preferred amino acids at this position are Leu and Pro for FheCL1 and FheCL3 , respectively , confirming our previous studies using short fluorogenic peptides [13] , [14] . The profile also shows that both enzymes share a strong selection against charged P2 residues ( Figure 2 ) . Also on the non-prime side , the juvenile enzyme , has a slight preference for Gly in the S3 pocket ( Figure 2 ) . This S3 preference is more noticeable at early digestion times ( 5 min . reaction ) , while other residues can be progressively accommodated at this site as the length of the incubation increases ( Figure S1 ) . On the prime side of FheCL3 , substrate preference is dominated by the P1′ site and shows a preference for Ser , Gly and to a lesser extent Met ( norleucine ) and Ala ( Figure 2 ) . Previous reports using internally quenched penta or heptapeptide substrates investigated the prime side preferences for papain and mammalian cathepsins B , L , S and K and showed that a broad range of amino acids were accommodated in these subsites [34] , [35] , [36] . However , while subtle differences were noted between the enzymes none of them can be considered as major contributions to specificity , except for a slight preference of hydrophobic moieties in papain P3′ [34] , and a general avoidance of Pro at P1′ [36] . Our data confirmed the avoidance of Pro , and highlights the preference of FheCL3 for Gly and Ser , a feature that might be relevant for the enzyme's ability to degrade collagen helical domains . Whereas the MSP-MS assay offers a more “natural” way of determining substrate specificity because the longer linear peptides are more like the loop regions found in protein substrates , Positional Scanning- Synthetic Combinatorial Libraries ( PS-SCL ) offer increased diversity for the study of P4 to P1 interactions since they comprise a collection of all possible fluorogenic tetrapeptides . This methodology has been widely used in the characterization of cysteine proteases [27] , [28] , and profiles of adult liver fluke proteases are known [20] , [29] . The PS-SCL profile for the recombinant FheCL1 used in this work is practically identical with that reported by Stack et al . [20] , independently supporting the accuracy of this tool in assessing enzyme specificity ( Fig . S1 ) . Furthermore , despite the differences in the methodological approaches , the PS-SCL results are generally consistent with the MSP-MS observations . FheCL1 displays a typical papain-like cysteine protease profile with S2 predominance , i . e . marked preferences for aliphatic residues , particularly Leu , at this position . Some minor selectivity can be found for the S1 interactions , where the basic residues Arg and Lys , together with Gln , Thr and Met are preferred . In contrast , the S3 and S4 pockets show a broad specificity completing a picture similar to that found by the MSP-MS analysis ( Figure 3 ) . The most obvious difference between FhCL1 and FheCL3 are the very distinct profiles observed for the P2 and P3 residues . The FheCL3 S2 pocket can accommodate Pro very readily , accepting it twice better than Val and four times better than Leu . In addition , unlike most known cysteine proteinases , the S3 pocket of FheCL3 demonstrates selectivity , specifically for Gly ( Figure 3 ) . Consistent with the PS-SCL data the MSP-MS results at 5 min of digestion shows FheCL3 has a preference for Gly in P3 ( Figure 2 ) , and as the reaction proceed other amino acids are also accommodated in S3 as indicated by an increased frequency in later times . This effect is expected since in the MSP-MS assay all peptides are mixed and assayed together , consequently the preferential cleavages would be observed early in the reaction . Selectivity at S3 is relatively rare , although PS-SCL studies have shown that the plant enzymes papain and bromelain have a noticeable preference for Pro at P3 [27] . Our previous studies showed that FheCL2 also has a slightly increased preference for Gly at P3 , and an augmented acceptance of Pro at P2 although maintaining Leu as the preferred residue in this position [20] . Therefore , FheCL2 active site appears to have intermediate characteristics between FheCL1 and FheCL3 , both at S2 and S3 subsites ( compare Figure 3 with that of Stack et al . [20] http://www . jbc . org/content/283/15/9896 . full . pdfhtml ) . The PS-SCL profile of FheCL5 , an enzyme secreted in very low abundance by adult F . hepatica , has also been reported and is more similar to FheCL1 with strong Leu preference at P2 , although it has the unique ability to accept Asp [29] . These results support the idea of functional divergence and specialization of the different members of the liver fluke cathepsin L family occurred following several gene duplications as proposed by phylogenetic analysis [5] , [6] . Since non-prime side differences in specificity between Fasciola cathepsin Ls are mainly restricted to S2 and S3 , we investigated the contribution of the variable residues lining those sites by mutation analysis . These pockets differ only at three positions: 61 , 67 and 205 located at the bottom of the S3 pocket , at the hinge of subsites S2 and S3 , and at the bottom of the S2 subsite , respectively ( Figure 4 ) . The first two variations involve amino acids with different properties , while the third involves a substitution between similar aliphatic moieties . Based on these observations , we changed residues 61 and 67 of FheCL3 , for those present in FheCL1 , generating the variants FheCL3 H61N , FheCL3 W67L and a double mutant bearing both substitutions . Their preferences at P2 and P3 were assessed with the PS-SCL approach . FheCL3 H61N showed only a subtle change in enzyme specificity , decreasing Gly preference in relation to the other amino acids as has been predicted ( Figure 5 B ) . The FheCL3 W67L variant resulted in a marked reduction in the preference for Pro at P2 compared to FheCL3 , while simultaneously increasing the aliphatic residues preferences and making Val the most favorably accommodated residue . Importantly , we found that the FheCL3 W67L variant also altered P3 specificities , changing the Gly preference to an increased preference for Leu ( Figure 5 C ) . The double mutant enzyme , FheCL3 H61N/W67L , presented S2 and S3 profiles similar to the single FheCL3 W67L change ( Figure 5 D ) . To complete the picture , we engineered the FheCL1 S2 pocket to resemble that of FheCL3 by replacing the key residues at positions 61 and 67 ( Figure 4 ) . Based on the PS-SCL neither of the changes introduced could modify FheCL1's preference for Leu at P2 , nor increase significantly its acceptance of Pro in that position ( Figure 6 ) . However , the substitution of Leu67 to Trp did slightly increased FheCL1's acceptance of Gly in P3 , either in the single change variant ( Figure 6 C ) or in the double mutant ( Figure 6 D ) . Furthermore , in these Trp-containing variants ( FheCL1 L67W , and FheCL1 N61H/L67W ) , the preference for Pro at P3 increase in comparison with the wild type enzyme ( Figure 6 C–D ) , suggesting that the change is restricting S3 to small residues . The N61H single change imparts minor effects on S3 selectivity , suggesting that the entrance and not the bottom of the S3 pocket is crucial for selectivity ( Figure 6 B ) . Taken together our data shows that a single change at position 67 is sufficient to strongly reduce the unique specificities of FheCL3 at both S2 and S3 sites , and moreover , rearrange the whole active site pockets contribution to substrate recognition . Modifications at this position in FheCL1 had little effect on substrate specificity . Therefore FheCL1's preference for Leu at P2 seems to be robust and does not depend only on the residues lining the S2 or S3 pocket that were evaluated in this work . Different effects of modifications at position 67 have already been reported , in mammalian cathepsins [25] , [37] and in the liver fluke proteases [29] , [38] but these studies in general did not analyzed the possible contributions of the residues occupying the S3 pockets . To support the data we observed with PS-SCL , we investigated the enzyme kinetics of the parent enzymes and their variants using two fluorogenic tripeptide substrates , Z-VLK-AMC and Tos-GPR-AMC , which are representative of the FheCL1 and FheCL3 subsite preferences . The calculated kinetic parameters KM , kcat and kcat/KM and the variation imposed by the diverse variants examined are presented in Table 1 . We found that substitutions made at the active site residue 67 of FheCL3 resulted in a marked reduction in enzyme efficiency for both substrates ( this was also seen with the PS-SCL ) . Compared to the parent enzyme , FheCL3 W67L exhibited a drastic diminution in specificity towards Tos-GPR-AMC ( 1440-fold ) , predominantly due to a major reduction in the catalytic turnover constant kcat . A less pronounced , though also large ( 35-fold ) decrease in specificity towards Z-VLK-AMC ( Table 1 ) was observed . The double variant FheCL3 H61N/W67L presented a profile very similar to the FheCL3 W67L single mutant , suggesting a minimal contribution from the H61 in the S3 subsite . When analyzing the variations in the S2 pocket of FheCL1 we found that the variant FheCL1 L67W showed a decrease in specificity for peptides with Leu in P2 ( Z-LR-AMC or Z-VLK-AMC ) . These were 8 times lower predominantly due to a decrease in the kcat of the modified enzyme . This substitution only slightly increased the activity of FheCL1 towards Tos-GPR-AMC and hence the FheCL1 L67W variant did not nearly approach the specificity observed by FheCL3 for this substrate ( Table 1 ) . The FheCL1 S3 pocket replacement , FheCL1 N61H ( like in FheCL3 ) did not alter the specificity of the enzyme towards Z-VLK-AMC and resulted in a slight increase ( 1 . 4-fold ) in its activity towards Tos-GPR-AMC , likely due to a better accommodation of Gly at P3 which would be consistent with the observations of the PS-SCL analysis ( Figure 6 ) . Consequently , despite finding the expected variations in Z-VLK-AMC and Tos-GPR-AMC activity in FhCL1mutants , these changes are not enough to absorb the more than 200-fold difference in specificity that FheCL1 has for these two types of substrates and the enzymes still prefer substrates with P2 Leu ( Table 1 ) . Previously , Stack et al . [20] found that the L67Y change in FheCL1 did not significantly modify the activity towards Tos-GPR-AMC which is consistent with our studies . However , a 13-fold increase on the activity towards this substrate was observed when a similar L67Y change was introduced into FheCL5 [38] . FheCL5 active site is more restricted at both the S2 and S3 pockets than FheCL1 due to the presence of the bulkier Leu157 and Tyr61 residues respectively . The L67Y change would impose a further restriction in the active site such that small residues at P3 and P2 would be favored . Consequently , the improved acceptance of Tos-GPR-AMC could be explained by the presence of the adjacent Gly and Pro positioned at P3 and P2 respectively , rather than by the modest rise in activity towards Pro at P2 as originally proposed [38] . The same rationale explains the recent observation that a FheCL5 L67F mutation increased activity towards Tos-GPR-AMC , and the inverse FheCL2 Y69L variant reduced P2 Pro acceptance [29] . Given the unusual characteristic of FheCL3 to efficiently degrade native type I collagen , we assessed the efficacy of the parent FheCL3 and its variants to hydrolyze type I collagen in vitro . Unlike wild type FheCL3 , both FheCL3W67L and FheCL3 H63N/W67L variants were unable to cleave collagen at neutral pH and 28°C , conditions that preserve its native structure ( Figure 7 ) . The reduced activity of FheCL3 mutants indicate that Trp67 might be crucial for the enzyme activity that might be centered in cleaving substrates enriched in small amino acids ( Gly , Pro ) like collagen . Our findings agree with previous observations that the substitutions Y67L and L205A in human cathepsin K ( for residues present in human cathepsin L ) , abolish its collagenolytic activity [37] . This human cathepsin K variant acquires the S2 preferences of human cathepsin L , and the reciprocal replacements to human cathepsin L conferred it with a specificity similar to cathepsin K [25] . We have also prepared a double variant of FheCL1 at the same positions , i . e . FheCL1 L67Y/L205A but this did not exhibit collagenolytic activity ( data not shown ) . This lack of correlation with human cathepsin L and K mutants behavior is surprising , although differences at other positions within the active sites exist between the mammalian and fluke enzymes that must also be important in determining collagenolytic ability . These differences in turn might prove useful in the design of specific inhibitors or drugs for the parasite enzymes over host homologues . Our analysis of active site variants highlights the role of residue 67 which is determining by its gate-keeper position not only the conformation of the S2 subsite , but also of the S3 pocket . Using molecular modeling we analyzed the possible conformations of Trp67 in the active site of FheCL3 as compared to FheCL1 ( Figure 8 ) . The most stable rotamer protrudes and partially occludes the S2 subsite ( Figure 8 B ) . An alternative conformer places the indole ring towards the S3 subsite reducing this site volume ( Figure 8 C ) , while a third low energy rotamer is coaxial with the active site cleft leaving two more open but narrow active site pockets ( Figure 8 D ) . The rotation of this residue might be fundamental to accommodating the distinct substrates of FheCL3 , defining the nature of the amino acids that can be accepted in these subsites . The planar ring of Pro occupying the P2 subsite can be stabilized by stacking interactions with the aromatic heterocycle of Trp . Furthermore , aliphatic moieties can also be accommodated at this site due to the hydropobic nature of FheCL3 S2 pocket . However , at the same time than stabilizing some interactions the bulky Trp can be imposing steric hindrances in the neighbor subsite thus favoring small residues . Based on this observation we reanalyzed the MSP-MS data looking at the amino acid pairs present at S3-S2 . We noticed that FhCL3 can accommodate different residues at P3 if P2 is occupied by Pro , and that tiny Gly is slightly preferred at early times combined either with Pro or aliphatic moieties . In fact if small residues are present in P3 , other residues can be placed in P2 excepting aromatic ones , which are disfavored in any combination by FheCL3 ( data not shown ) . These combined preferences for Pro and to a lesser extent for Gly residues by FheCL3 , can explain why native collagen , that is rich in these amino acids is an appropriate substrate for this enzyme . We have characterized the FheCL3 cysteine protease of the infective larval stage of F . hepatica that exhibits a particular collagenolytic activity and analyzed the differential contribution of active site residues involved . Our results highlight that a Trp residue strategically located at the gatekeeper position between the S2 and S3 active site pockets is vital to this activity and contributes to narrow and constrained pockets that can best accommodate small residues , particularly , Pro at P2 and Gly at P3 . These peculiarities are not shared by other known cysteine proteases , suggesting that the enzyme may be a good target for the development of small molecule inhibitors for parasite control . Furthermore , our mutation analyses reveal the under-appreciated significance of interactions at P3 that together with those at P2 contribute to modulating cysteine protease specificity . Novel extended peptide libraries provide first glimpses of other interactions particularly at the prime side of the active site cleft , showing noticeable differences whose contributions to specificity and selectivity need to be assessed in future studies .
|
The flatworm Fasciola hepatica is responsible for fasciolosis , one of the most common parasitic diseases of livestock worldwide , with increased incidence of human cases . When contaminated plants are ingested , infective larvae are released and transverse the gut wall before migrating to the bile ducts within the liver . Migrating liver flukes erode host tissue while adults feed on blood and they mature and release thousands of eggs . Several developmentally-regulated cathepsin L like proteolytic enzymes ( FheCLs ) are essential to the migrating and feeding processes . Despite being similar in structure and sequence these enzymes show specialization attacking preferentially different substrates and taking part in the diverse process of invasion , immune evasion and feeding . Our analyses reveal unique differences in activity between the major infective juvenile ( FheCL3 ) and adult ( FheCL1 ) enzymes , and demonstrate that the juvenile enzyme has a particular active site that allows it to degrade collagen , the main component of connective tissues . We demonstrate that a single position on the active site , residue 67 , is essential to this collagenolytic activity critical for parasite invasion .
|
[
"Abstract",
"Introduction",
"Methods",
"Results/Discussion"
] |
[
"biomacromolecule-ligand",
"interactions",
"medicine",
"biochemistry",
"infectious",
"diseases",
"fasciolosis",
"enzymes",
"neglected",
"tropical",
"diseases",
"biology",
"enzyme",
"kinetics",
"parasitic",
"diseases"
] |
2013
|
Dissecting the Active Site of the Collagenolytic Cathepsin L3 Protease of the Invasive Stage of Fasciola hepatica
|
Incompatibilities between the nucleus and the cytoplasm of sufficiently distant species result in developmental arrest of hybrid and nucleocytoplasmic hybrid ( cybrid ) embryos . Several hypotheses have been proposed to explain their lethality , including problems in embryonic genome activation ( EGA ) and/or nucleo-mitochondrial interactions . However , conclusive identification of the causes underlying developmental defects of cybrid embryos is still lacking . We show here that while over 80% of both Xenopus laevis and Xenopus ( Silurana ) tropicalis same-species androgenetic haploids develop to the swimming tadpole stage , the androgenetic cybrids formed by the combination of X . laevis egg cytoplasm and X . tropicalis sperm nucleus invariably fail to gastrulate properly and never reach the swimming tadpole stage . In spite of this arrest , these cybrids show quantitatively normal EGA and energy levels at the stage where their initial gastrulation defects are manifested . The nucleocytoplasmic incompatibility between these two species instead results from a combination of factors , including a reduced emission of induction signal from the vegetal half , a decreased sensitivity of animal cells to induction signals , and differences in a key embryonic protein ( Xbra ) concentration between the two species , together leading to inefficient induction and defective convergence-extension during gastrulation . Indeed , increased exposure to induction signals and/or Xbra signalling partially rescues the induction response in animal explants and whole cybrid embryos . Altogether , our study demonstrates that the egg cytoplasm of one species may not support the development promoted by the nucleus of another species , even if this nucleus does not interfere with the cytoplasmic/maternal functions of the egg , while the egg cytoplasm is also capable of activating the genome of that nucleus . Instead , our results provide evidence that inefficient signalling and differences in the concentrations of key proteins between species lead to developmental defects in cybrids . Finally , they show that the incompatibilities of cybrids can be corrected by appropriate treatments .
Investigation of the mechanisms generating the characters or phenotypes during development has revealed the importance of the nucleus and its DNA content in directing developmental processes [1]–[4] . Nonetheless , the cytoplasm of the egg is responsible for the specification of many early aspects of development , including polarity , as well as cleavage type and developmental timing [3]–[6] . In principle , for the nucleus of one species to be compatible with the cytoplasm of the egg of another species , the foreign species' nucleus must not interfere beyond a certain threshold with the maternally regulated developmental processes of the cytoplasmic ( egg ) species , while the recipient egg cytoplasm also needs to fully “activate” and support development promoted by the foreign nucleus . Thus , as a general rule , it is possible to generate viable offspring via interspecies Somatic Cell Nuclear Transfer ( iSCNT ) only if the egg cytoplasm and the donor nucleus come from two very closely related species or from sub-species , which develop in a highly similar manner . Indeed , if the two species used are sufficiently distant , the resulting embryos rarely progress normally through embryonic development and often arrest at the stage of EGA or soon after [2] , [3] , [7] , [8] . One of the first scientists who became interested in this field was Baltzer , who revealed the importance of the nucleus in development using androgenetic Triturus cybrids . Indeed , he observed that when a sperm from one species is combined with the egg cytoplasm of another species , androgenetic development differs from that when the sperm is of the same species as the egg , and in fact leads to severe developmental defects [1] . He , and others , further recorded differences between the development of reciprocal androgenetic cybrids [9]–[11] , which could in principle suggest that the basis of the incompatibilities between the respective nuclei and cytoplasms of two given species might not necessarily be reciprocal . Later , the method of nuclear transfer [12] not only enabled the transplantation of diploid nuclei into enucleated eggs in virtually any species combinations , but also allowed the transfer of nuclei from cybrid embryos back to their own species egg cytoplasm . Using this technique , Moore ( 1958 ) showed that the nucleocytoplasmic incompatibilities between two Rana species ( R . pipiens and R . sylvatica ) led to irreversible nuclear damage [13] . Similar conclusions were later attained when back-transfer experiments were performed with the cybrids made from two Xenopus species ( X . laevis and X . tropicalis ) , suggesting that irreversible nuclear damage may be a common effect of nucleocytoplasmic incompatibilities [14] . Interestingly , cybrid lethality was shown to occur even in a combination ( R . palustris nuclei into R . pipiens cytoplasm ) in which no cytologically detectable chromosome damage was found to occur and back-transferred embryos developed normally [15]–[17] , suggesting that nuclear damage is not the whole explanation for developmental defects in cybrids . Also , a few experiments in which pieces of cybrid embryos were grafted onto normal embryos of either parental species suggested that the developmental defects of these embryos were cell autonomous , as contact with normal tissue did not rescue their developmental potentials [14] , [16] . Finally , a more extreme cybrid combination ( D . pictus nucleus into X . laevis egg cytoplasm ) also generated by iSCNT , arrested before gastrulation , had reduced mRNA synthesis and did not initiate rRNA synthesis [18] . It , however , remains unclear whether these defects were the primary causes of the arrest , or secondary to other incompatibilities . Work later performed in fish using iSCNT suggested that major differences in chromosome numbers could be one of the essential factors causing the nucleocytoplasmic incompatibilities [19] , [20] . Interestingly , chromosome loss was observed in lethal fish hybrids generated by cross-fertilization , while in one such combination , phospho-histone H3 abnormally persisted on the lagging chromosomes during anaphase [21] , [22] . Consistent with the amphibian work , cybrid lethality can , however , occur without any obvious defects in chromosome segregation , since a fish cybrid combination ( goldfish nucleus into loach egg cytoplasm ) that does not suffer from chromosome elimination is embryonic lethal [20] , [23] . In the meantime , massive experimentation with iSCNT in mammals has also explored the limits of this technique and provided new insights regarding the potential causes of the developmental arrest in lethal cybrid combinations . One main conclusion derived from several reports suggested that a major barrier to cybrid development must be manifested at the stage of EGA , since it coincides with the stage of arrest of a majority of lethal mammalian cybrid combinations [7] , [8] . This hypothesis was recently supported by transcriptional analyses [24]–[27] . A second possibility is a potential incompatibility between the maternal mitochondrial genome and that of the foreign species nucleus , leading to defects in mitochondrial function in cybrids [7] . A reason for this suspicion comes from the fact that higher mutation rates ( compared to that of genomic DNA ) combined with maternal inheritance can lead to rapid divergence in mitochondrial DNA during evolution [28] . Also , the efficiency of same-species bovine SCNT is increased if the donor and recipient cells have the same mitochondrial haplotype [29] , while ATP levels were reduced in chimpanzee/bovine iSCNT embryos [27] . Finally , evolutionary distances as little as 8–18 million years lead to fatal defects in oxidative phosphorylation in primate and rodent xenomitochondrial cybrid cell lines [30] , . Pioneering work has thus established that there are developmental incompatibilities between the nucleus and the cytoplasm of sufficiently distant species . Yet the rules that determine the compatibility between the maternal cytoplasmic content and that of the nucleus in the context of early development remain poorly defined . Equally obscure are the precise initial faults in the developmental mechanisms that eventually lead to the arrest in cybrid embryos . To better understand the nature of the nucleocytoplasmic incompatibilities that exist between relatively distant species , we have analysed here the developmental potentials and defects of reciprocal X . laevis and X . tropicalis hybrids and those of the lethal cybrids formed by the combination of a haploid X . tropicalis sperm nucleus and a X . laevis egg cytoplasm .
In all haploid Anura , the onset of gastrulation is delayed by the time it takes for all the cells to undergo approximately one additional division , until they reach the same nucleocytoplasmic volume ratio as in diploid embryos [34] . In addition to this developmental retardation , X . laevis haploid embryos are microcephalic , have a shorter axis , and suffer from lordosis and a bulging abdomen ( Figure 1A–B; [35] ) . Haploids also have a feeble heart , are much less active than diploids , and are subject to oedema , such that no haploid X . laevis has ever reached metamorphosis [34] , [35] . Consistent with this , [l]xl embryos showed all of the early phenotypes described above ( Figure 1A–B , Table 1 , Videos S1–S2 ) . Development of lx[l] embryos was also briefly investigated and found to be identical to that of [l]xl embryos ( unpublished data ) . We further found that the development of [t]xt embryos was comparable to that of [l]xl or lx[l] in all respects ( Figure 1C–D , Table 1 ) . Therefore , both species possess a roughly equal early developmental potential in the androgenetic haploid state , developing into similarly advanced stunted swimming tadpoles with a frequency above 80% ( Figure 1A–D , Table 1 ) . We then asked whether the addition of a haploid nucleus from one of these species would interfere with haploid development of the other species . Reciprocal in vitro cross-fertilization between X . laevis and X . tropicalis has previously been reported [36]–[38] . Whereas the cross-fertilization of the eggs of X . tropicalis with X . laevis sperm is very inefficient ( ∼3% ) , that of X . laevis eggs with X . tropicalis sperm is comparable to that of X . laevis self-fertilization ( unpublished data; [37] ) . It has been stated that lxt hybrids are viable and can develop to the adult stage [36] , yet further characterization was lacking . We have thus analysed the development of the reciprocal hybrids that can be generated from X . laevis and X . tropicalis . The lxt hybrid embryos develop into swimming tadpoles ( stage 40 ) with a frequency that is comparable to that of lxl or txt control embryos ( Figure 1E , Table 1 , Text S1 , Figure S1 ) , suggesting that the addition of a haploid set of X . tropicalis chromosomes does not interfere with gynogenetic haploid X . laevis development . Interestingly , the defects of lxt hybrids are highly reminiscent of those seen in X . laevis haploids ( [l]xl or lx[l] ) , although they have a markedly decreased severity in the hybrids ( Figure 1A–B , E , Table 1 ) . Thus , an additional haploid set of X . tropicalis chromosomes is in fact beneficial to X . laevis haploid development since lxt hybrids develop further than X . laevis haploids ( [l]xl or lx[l] ) , which never reach metamorphosis ( Figures 1A–E , S1 , Table 1; [34] , [35] ) . The reciprocal txl hybrid embryos all died as late blastulae or very early gastrulae ( Table 1 ) . This indicates that an additional haploid set of X . laevis chromosomes is very damaging to the early development of a gynogenetic X . tropicalis haploid . We have not further characterized this arrest . The results thus suggest that a haploid X . tropicalis nucleus does not interfere with X . laevis maternal and embryonic development , while a haploid X . laevis nucleus severely interferes with X . tropicalis maternal or embryonic developmental processes . Since a haploid X . tropicalis nucleus did not interfere with X . laevis haploid development , we next asked whether the cytoplasm of a X . laevis egg can support the normal development promoted by a X . tropicalis nucleus . Earlier experimentation with iSCNT suggested that the cytoplasm of the X . laevis egg was not capable of reprogramming and/or sustaining normal development promoted by a X . tropicalis neurula/tailbud stage somatic nucleus [14] . If this nucleocytoplasmic incompatibility between the X . laevis egg cytoplasm and a X . tropicalis somatic nucleus was not due to defects linked to nuclear transfer or reprogramming , the cytoplasm of a X . laevis egg should also be incapable of sustaining the normal development promoted by a X . tropicalis sperm nucleus to the swimming tadpole stage . We therefore compared the development of [l]xt cybrids to that of [l]xl or [t]xt same-species controls . The [l]xt cybrids developed relatively normally at first and were indistinguishable from control [l]xl androgenetic haploids until they reached the beginning of gastrulation and the appearance of the dorsal lip of the blastopore ( stage 10 . 25 ) , approximately 1 h after lxl diploids ( Table 1 , Videos S1–S2 ) . However , the [l]xt cybrid embryos subsequently showed developmental retardation and consistently failed to close their blastopore , formed abnormal neurulae , and all died as abnormal , non-swimming postneurulae ( Figure 1F–G , Table 1 , Video S2 ) . The proportion of embryos reaching a postneurula stage ranged from <10% to >80% depending on male/female combinations ( or egg batches ) . The most developmentally advanced of these cybrid embryos had a rudimentary sucker , microcephalic head , and pigmented elementary eyes . A few ( <5% ) also sporadically underwent bursts of rhythmic muscular contractions and/or developed a primitive caudal fin , and very few ( <1% ) also showed posterior axis elongation ( Figure 1G shows the most developed [l]xt embryo obtained , while Figure 1F shows a more typical example ) . Exceptionally well-developed [l]xt cybrid individuals could survive for up to almost a week . Overall , the terminal phenotype of these [l]xt cybrid embryos is very similar to those ( diploids ) previously obtained by iSCNT [14] . One difference , however , is that they are less elongated , which is likely to be the result of the difference in ploidy , since same-species Xenopus haploids are readily characterized by reduced axis elongation ( Figure 1A–D; [34] , [35] ) . Thus , [l]xt androgenetic haploid cybrids have a reduced developmental potential compared to same-species androgenetic haploid controls ( [l]xl or [t]xt ) , demonstrating the existence of a developmental nucleocytoplasmic incompatibility between these two species that is not due to nuclear transfer or reprogramming defects . Even though the replication of X . tropicalis nuclei in X . laevis cytoplasm may trigger unknown nuclear aberrations [14] , chromosome loss was not observed in the [l]xt cybrid embryos ( 4/4 [l]xt embryos had cells in which the expected haploid chromosome complement of X . tropicalis was clearly visible in metaphase squash preparations; unpublished data ) . Attempts to generate the reciprocal [t]xl androgenetic haploid cybrid were unsuccessful , probably owing to the low efficiency of cross-fertilization in this direction ( Table 1 ) , and thus the developmental potential of a haploid X . laevis nucleus in a X . tropicalis egg cytoplasm remains undefined . These results thus indicate that even though the presence of a X . tropicalis haploid nucleus does not interfere with ( and even improves ) gynogenetic X . laevis development , the X . laevis egg cytoplasm does not support the development that is promoted by a X . tropicalis nucleus as well as the X . tropicalis egg cytoplasm . Furthermore , it establishes the onset of gastrulation ( stage 10 ) as the critical stage where the nucleocytoplasmic incompatibility is first manifested . A major barrier to the development of cybrid embryos is believed to reside at the stage of EGA [7] , [18] , [24]–[27] . Suppression of transcription with α-amanitin ( intra-cytoplasmic concentration of 50 µg/ml ) causes X . laevis embryos to arrest prior to gastrulation ( unpublished data; [39] ) . It is therefore plausible that the components of the X . laevis egg cytoplasm are unable to efficiently activate transcription from the X . tropicalis genome , resulting in the observed gastrulation defects . To test this , we used quantitative RT-PCR to evaluate the mRNA content of several embryonically transcribed genes in stage 10 . 25 [l]xt cybrid embryos . The relative quantity of transcripts for Xbra , Chordin , Gata4 , and Mixer at this stage in [l]xt cybrids was not significantly different from txt or [t]xt control embryos ( Figure 2A ) . We therefore conclude that the gastrulation defects of [l]xt cybrids do not arise from a generalized inefficient EGA . To exclude the possibility that differences in the splicing or translation machineries between the two species could instead lead to inefficient protein synthesis from these properly concentrated embryonic transcripts following EGA , we investigated the production of Xbra protein in [l]xt cybrids . Western blot comparison of stage 11 embryos revealed that the relative concentration of Xbra protein in [l]xt cybrid embryos was similar to that in control X . laevis embryos ( lxl and [l]xl ) ( Figure 2B ) . It may be important to note that Xbra protein concentration is markedly reduced in X . laevis egg-based embryos ( lxl , [l]xl , [l]xt ) relative to X . tropicalis egg-based embryos ( txt , [t]xt ) ( Figure 2B–C ) . This suggests that the concentration of Xbra protein at stage 11 is different in X . laevis and X . tropicalis embryos , and maternally/cytoplasmically regulated in the cybrid . However , since the level of Xbra protein present in the X . laevis egg-based embryos is similar regardless of whether it is encoded by a X . laevis or X . tropicalis genome , these results suggest that the early gastrulation defects of the cybrid embryos do not result from a generalized deficiency in EGA or protein synthesis . The last phase of EGA in Xenopus consists of the activation of rDNA transcription and nucleologenesis [40] , [41] . Nucleologenesis requires factors present in the oocyte nucleolus in mammalian embryos [42] , and was defective in cybrids generated by iSCNT [18] , [26] , [43] . To verify whether the last phase of EGA is completed in [l]xt cybrids , we analysed nucleologenesis ( the nucleolus itself results from active rDNA transcription [44] ) in these embryos . No statistical difference was , however , found regarding nucleoli numbers between the nuclei of the [l]xt cybrids and control haploids ( [l]xl or [t]xt ) ( Figure S2A–G ) , indicating that the X . laevis cytoplasm efficiently recognizes the X . tropicalis nucleolar organizer . Nucleolar integrity in [l]xt cybrids was further confirmed by the broad intra-nucleolar distribution of fibrillarin [26] , identical to the controls ( Figure S2H–J ) . Interestingly , lxt hybrid embryos had significantly fewer nuclei with two nucleoli than either diploid controls ( lxl and txt ) ( Figure S2A–G ) , suggesting that one of the nucleolar organizers is partially dominant . Our results , however , suggest that nucleologenesis , and thereby rRNA synthesis , is successful and that EGA is therefore completed in [l]xt cybrids; their early gastrulation defects must therefore arise from other incompatibilities . An incompatibility between the maternal species mitochondrial genome and the foreign species nuclear-encoded mitochondrial genes could lead to deficient energy production and underlie lethality in cybrids [7] , [45] . Mitochondrial ATP synthesis is indeed required for Xenopus embryos to initiate gastrulation ( 100% of X . laevis or X . tropicalis embryos ( n = 30 each ) arrested at stage 9 when cultured in 40 µM oligomycin ( unpublished data ) , an inhibitor of mitochondrial ATP synthase [46] ) . We used a luciferase-based assay to determine the absolute ATP content at various time points during early embryonic development in the diverse kinds of X . laevis egg-based embryos ( lxl , [l]xl , lxt , and [l]xt ) . The average number of ATP molecules per X . laevis egg obtained by this method was 1 . 5 nmoles ( from three different frogs ) , close to the 1 . 6 nmoles for in vitro matured oocytes that was previously measured using chromatography [47] . Overall , the ATP content in all X . laevis egg-based embryos tested decreased until stage 10 . 25 to about 2/3 of the egg content , and then remained constant or slightly increased until stage 11 . 5 ( Figure 3A ) . The ATP content curves of the two kinds of diploid embryos ( lxl and lxt ) were very similar to each other , while those of the two kinds of haploid embryos ( [l]xl and [l]xt ) appeared slightly different from the diploid curves , which may reflect different energy dynamics between haploid and diploid embryos . No statistical difference ( p>0 . 05; two-tailed t test ) in ATP content was found between [l]xt cybrids and control [l]xl sibling embryos at any time point until stage 11 . 5 ( Figure 3A ) . Thus , we conclude that the early gastrulation defects in [l]xt cybrid embryos are not due to reduced ATP levels . Energy stress in all eukaryotes is detected in a very sensitive manner by the AMP-activated protein kinase ( AMPK ) , which becomes phosphorylated in its activation loop following an increase in the AMP∶ATP ratio [48] . We have thus used anti-phospho-AMPK antibodies to detect AMPK phosphorylation in various kinds of embryos at stage 11 , well after the onset of the gastrulation defects of [l]xt cybrids . The level of AMPK phosphorylation in these cybrids was similar to that of control embryos ( Figure 3B ) . Therefore , we conclude that the gastrulation defects in [l]xt cybrid embryos are not due to ATP depletion or energy stress . It appears unlikely that an incompatibility between the X . laevis mitochondria and the X . tropicalis nucleus that would not affect ATP levels could explain the early gastrulation defects occurring in these cybrids . To gain insights into the mechanisms responsible for the developmental faults of [l]xt cybrid embryos , we sought to understand the basis of their early gastrulation defect , namely the failure to close their blastopore and elongate their body axis ( Figure 1 , Video S2 ) . Blastopore closure and body axis elongation are both highly dependent on efficient convergence and extension of the involuting marginal zone [49] . To test the efficiency of induction and convergence-extension movements in the gastrulating cybrid embryos , we compared the elongation of stage 10 . 5 dorso-marginal explants from these embryos to that of similar explants from control embryos . We adopted the following system to score the induction response [50] . If the explants are not induced , they remain spherical ( no elongation ) . If induction and efficient convergence-extension occur , the explants elongate such that their length/width ratio becomes greater than two ( well elongated ) . If the explants are induced but do not undergo efficient convergence-extension , they only partially elongate ( stump ) . Over 70% of stage 10 . 5 dorso-marginal explants taken from control embryos ( lxl , [l]xl , txt , or [t]xt ) underwent efficient convergence-extension , while the remaining also elongated , but to a lesser extent ( Figure 4 , Table 2 ) . In contrast , few ( 14% ) of the explants from [l]xt cybrid embryos underwent efficient convergence-extension , while most ( 67% ) elongated to a lesser extent and some ( 19% ) did not elongate ( Figure 4 , Table 2 ) . Therefore , we conclude that the dorso-marginal region of [l]xt cybrid embryos is defective in induction response and convergence-extension during gastrulation , and this may be responsible , at least in part , for their incapacity to close their blastopore and properly elongate their body axis . Gastrulation movements and convergence-extension in Xenopus are driven by cells of the mesoderm , which arises at the equatorial region following the perception of a mesoderm-inducing signal that is generated by the vegetal hemisphere . If the cells of the animal hemisphere are not exposed to this signal , they remain ectodermal and do not elongate [51] , [52] . Therefore , the reduced elongation response of the cells originating from the animal hemisphere in [l]xt cybrid embryos could in principle result either from deficient induction signal emission from the vegetal cells , or from a defective response of the animal cells to correct levels of induction signals , or both . To determine whether the vegetal hemisphere of the [l]xt cybrids secrete signals capable of inducing efficient convergence-extension in adjacent animal cells , we compared the elongation of naïve lxl animal caps ( stages 8–9 ) that were conjugated to same-stage vegetal hemispheres of the following kinds: lxl , txt , [t]xt , and [l]xt . Whereas the vegetal halves of control embryos ( lxl , txt , and [t]xt ) were equally good at inducing lxl animal cap elongation , there was a marked reduction in the proportion of animal caps efficiently elongating following induction by the vegetal hemisphere of [l]xt cybrids ( Figure 5 , Table 3 ) , suggesting that reduced emission of inductive signals by the vegetal half of the cybrid embryos may contribute to their convergence-extension defects . Nonetheless , a significant proportion ( 30% ) of the lxl animal caps that were conjugated to [l]xt cybrid vegetal halves demonstrated efficient elongation , identical to the controls , suggesting that the vegetal cells of [l]xt cybrids can provide sufficient mesoderm-inducing signals to trigger efficient convergence-extension and elongation of animal cap cells , albeit in a reduced proportion of embryos ( Figure 5 , Table 3 ) . It seems therefore unlikely that the only problem underlying the gastrulation defects , which occur in 100% of [l]xt cybrid embryos , is a deficient secretion of mesoderm-inducing signals by their vegetal hemisphere , although this may indeed contribute to the problem in many embryos . We thus investigated the possibility that the cells of the animal hemisphere in [l]xt cybrid embryos do not respond properly , even to normal levels of mesoderm-inducing signals . We compared the response of unspecified animal caps ( stages 8–9 ) isolated from [l]xt cybrid embryos that were conjugated to diverse kinds of same-stage vegetal hemispheres ( lxl , txt , [t]xt , and [l]xt ) . Strikingly , elongation of these animal caps was only marginally ( ∼10%–20% ) improved by their conjugation to any of the different non-cybrid vegetal halves tested ( lxl , txt , [t]xt ) ( Figure 5 , Table 3 ) . This suggests that the animal cap cells in the majority of the cybrid embryos do not undergo efficient convergence-extension , even if exposed to normal levels of mesoderm-inducing signals , coming from the vegetal hemispheres of either species' embryos . In contrast , the elongation of [t]xt animal caps in this assay was not significantly different from that of lxl ( Table 3 ) , confirming that the poor elongation of [l]xt animal caps is not solely the result of their ploidy . Therefore , the reduced elongation response of animal cap cells of [l]xt cybrid embryos results both from deficient induction signal emission from their vegetal hemisphere and from a defective response of their animal cells , even to a normal level of inductive signals . Mesoderm specification and animal cap elongation can be induced in vitro in a dish containing nanomolar concentrations of Activin A in a dose-dependent manner [51] , [52] . In such an assay , animal caps isolated from X . laevis or X . tropicalis diploid embryos both have the same competence to respond to activin in terms of the induction of differentiation and gene expression [53] . We used this system to further test the induction and elongation efficiency of stage 8 animal cap cells isolated from [l]xt cybrid embryos . As expected , a significant proportion of the animal caps isolated from control embryos ( lxl , [l]xl , txt , and [t]xt ) elongated well in response to activin in a dose-dependent manner ( 5 ng/ml for 20 or 60 min ) , although the elongation was generally less efficient in haploid embryos ( Figure 6 , Table 4 ) . This was expected since axis elongation in haploid embryos is reduced compared to diploids ( Figure 1; [34] , [35] ) . A reduced proportion of naïve animal caps isolated from [l]xt cybrid embryos elongated in response to similar doses of activin in a dose-dependent manner , but strikingly they never underwent efficient convergence-extension ( Figure 6 , Table 4 ) . This result confirms that the reduced convergence-extension in the cybrid embryos largely results from a deficient response of the animal cap cells , even to normal levels of mesoderm-inducing signals . If the sensitivity to activin is compromised in [l]xt cybrid embryos , further increasing the activin induction treatment might be expected to rescue their defects in induction response and convergence-extension . Increasing the activin treatment , either by quintupling activin concentration or doubling the treatment time , indeed caused a higher proportion of the animal caps to elongate , and a few even underwent efficient convergence-extension ( Figure 6 , Table 4 ) . These results indicate that the sensitivity to activin is compromised in [l]xt cybrid embryos . However , even if the induction treatment is increased to ensure the perception of induction signals ( and an elongation response ) in almost all embryos , the vast majority of these still do not undergo efficient convergence-extension ( Figure 6 , Table 4 ) , suggesting that other incompatibilities manifest themselves by preventing efficient convergence-extension to occur during gastrulation . We observed that Xbra protein concentration is markedly lower in all X . laevis egg-based embryos ( lxl , [l]xl , [l]xt ) compared to X . tropicalis egg-based embryos ( txt , [t]xt ) ( Figure 2B–C ) . Following the induction of mesodermal cells , one function of Xbra consists of suppressing migratory movements to instead promote convergence-extension [54]–[56] . One possibility is therefore that a lower ( X . laevis–like ) concentration of Xbra protein does not suppress cell migration enough to permit convergence-extension in cells with a X . tropicalis genome . To test this hypothesis , we overexpressed Xbra in [l]xt cybrid animal caps prior to activin treatment . As expected , such treatment did not affect the proportion of cybrid animal caps that responded to activin induction by undergoing some degree of elongation , while a few of them underwent efficient convergence-extension ( Table 4 ) . When combined with prolonged activin exposure , this treatment rescued convergence-extension in 29% of cybrid animal caps ( Figure 6 , Table 4 ) . These results together suggest that the maternally regulated difference in Xbra protein concentration between the two species is partly responsible for the inefficient convergence-extension in gastrulating [l]xt cybrids , while reduced mesoderm-inducing signal emission and sensitivity also contributes to this phenotype . To validate these conclusions , we have attempted convergence-extension rescue in whole cybrid embryos using means expected to upregulate induction and/or Xbra signalling . One consisted in the injection of Activin A protein into the blastocoel of [l]xt cybrid blastulae , and the second in widely overexpressing FRL-1 , an EGF-CFC family member that is a limiting co-factor in nodal signalling and mesoderm induction [57] , [58] . These treatments both significantly improved blastopore closure and embryo elongation ( Figure 7 ) , two processes whose success is highly dependent on efficient convergence-extension . Widely overexpressing Xbra in whole cybrid embryos also improved elongation ( p = 4×10−7 ) , but it impaired blastopore closure and the resulting embryos were highly abnormal ( unpublished data ) . These results further support the hypothesis that the nucleocytoplasmic incompatibilities that lead to inefficient convergence-extension in [l]xt cybrid embryos result from deficient induction signalling and response , and from inadequate Xbra protein concentration .
It is estimated that X . laevis and X . tropicalis diverged from a common ancestor approximately 50–65 million years ( MY ) ago [59]–[62] . In comparison , humans are separated from the common chimpanzee by only approximately 6 MY [30] , while the extant placental mammal lineage evolved over approximately 135 MY [63] , [64] . Considering previous iSCNT reports of EGA defects in many lethal cybrids [18] , [24]–[27] , it was somewhat surprising to observe normal activation of key embryonic genes in our amphibian cybrid . However , the evolutionary separation between the mammalian and amphibian combinations tested in these articles is more considerable , respectively , ranging from ∼65 to ∼100 MY , and ∼235 MY [61] , [65] . Also , given that in X . laevis and X . tropicalis the expression patterns of transgenes generated with promoters from one species are generally maintained in the other species [66] , and that there is a high level of amino acid identity ( >98% overall ) between the homologous proteins of each species [67] , this result was not entirely unexpected . Moreover , because interspecies differences in transcription factor binding and gene expression are primarily directed by genetic sequence rather than cellular components [68] , in our case X . laevis transcription factors are expected to bind to and promote transcription of X . tropicalis genes in a X . tropicalis specific manner . Our data also corroborate a study in a lethal loach–goldfish cybrid in which early embryonic expression of two genes ( ntl and gsc ) took place normally [23] . Since only a handful of genes were tested in both cases , it remains possible that the embryonic transcription of other genes is aberrant in these “EGA-successful” cybrid embryos , and genome-wide transcriptional analyses will be necessary to address this question . One such study recently reported that rhesus-bovine iSCNT embryos activated hundreds of genes at EGA to levels comparable with in vitro fertilized rhesus embryos [69] . Defects in rRNA synthesis and nucleologenesis ( last step of EGA ) have been noticed in two distantly related cybrids , including those generated by the iSCNT of D . pictus nuclei into X . laevis eggs [18] , [26] . In both of these reports , however , mRNA synthesis ( first step of EGA ) was also substantially reduced , and thus it is conceivable that rRNA synthesis and nucleologenesis defects occurred as a secondary effect of the reduced mRNA synthesis rather than from an interspecies incompatibility . Alternatively , it is possible that the increased evolutionary distances between the combinations studied in these reports induced two independent incompatibilities leading to defects in both mRNA and rRNA synthesis . We did not observe any apparent nucleologenesis defects in our Xenopus cybrid . Also , since the terminal phenotype of anucleolate , rRNA-synthesis deficient , X . laevis mutant embryos [70] , [71] is considerably less severe than that of X . laevis egg-based cybrids [14] , [18] , it seems rather unlikely that potential defects in nucleologenesis and rRNA synthesis could be responsible for , or even contribute to , their observed developmental defects . Cybrid lethality may therefore occur even if the egg cytoplasm is able to properly activate a foreign nucleus' genome , such that the last stage of EGA is completed . The incompatibility of [l]xt cybrids does not result from interference of the foreign nucleus on the maternally programmed early developmental processes because the presence of a X . tropicalis sperm nucleus does not impair ( but improves ) gynogenetic X . laevis haploid development . We therefore conclude that cybrid lethality may occur even if the donor nucleus does not interfere with the recipient cytoplasm-regulated development , while the latter competently “activates” the donor species' nucleus . That we found no evidence of ATP deficiency or energy stress in [l]xt cybrid embryos was also somewhat unexpected , since it was shown that the combination of nuclei and mitochondria from cell lines of less distantly related mammalian species results in fatal defects in oxidative respiration [30] , [31] . In fact , our analysis does not exclude the possibility that ATP/energy-related nucleo-mitochondrial incompatibilities exist between X . laevis and X . tropicalis . If they do exist , however , they are not manifested early enough to generate energy stress in time to be responsible for the initial developmental defects that are apparent in [l]xt cybrid embryos . Alternatively , it remains possible that other , ATP/energy-unrelated , nucleo-mitochondrial incompatibilities could contribute to the gastrulation defects of the cybrids . We provide evidence that gastrulating cybrid embryos cannot execute efficient convergence-extension movements because of reduced levels of mesoderm-inducing signal emission by the vegetal pole of some of these , but also largely due to a defective elongation response of animal cap cells , in most if not all embryos . Exactly why this is taking place in the cybrids remains unclear , yet elongation of the animal cap cells of the cybrids can be partially rescued in explants and whole embryos by increasing exposure to a mesoderm-inducing signal ( Activin ) . One could hypothesize that the difference in egg sizes between the two species could affect induction signalling , or gastrulation movements . This , however , seems unlikely since our experiments clearly showed that the elongation defects of [l]xt cybrids occurred even in explants in vitro . Furthermore , the development of reciprocal inter-subspecies cybrids from X . laevis laevis and X . laevis victorianus was completely normal , despite over a 3-fold difference in the volume of their eggs [72] . We further show that the concentration of a key embryonic protein ( Xbra ) is different in the embryos of the two Xenopus species , while this concentration appears to be under cytoplasmic ( maternal ) control in cybrids . Since reduced Xbra activity inhibits convergence-extension in X . laevis [54]–[56] , and since Xbra overexpression partially rescues this defect in cybrids , it suggests that differences in key protein concentrations between species constitutes a form of nucleocytoplasmic incompatibility that contributes to developmental defects and lethality in cybrids . This phenomenon is likely not restricted to Xbra , since β-actin concentration is also markedly increased in X . tropicalis eggs and embryos compared to X . laevis ( unpublished data ) , while it was recently shown that different concentrations of Importin α and Ntf2 are responsible for the divergent nuclear sizes between these two species [67] . The mechanisms regulating protein concentration are largely unexplored , but from our analysis it appears that embryonic protein concentrations are under cytoplasmic ( maternal ) control , and not merely a reflection of mRNA concentrations , while inappropriate concentrations of key proteins in [l]xt cybrids , such as Xbra and likely others , may underlie their observed developmental defects and lethality . Quantitative proteomics analysis in [l]xt and other kinds of cybrids should reveal the magnitude of this mechanism . Finally , our ( and others [73] ) results demonstrate that it is possible to correct nucleocytoplasmic incompatibilities of cybrids by appropriate treatments . If nuclear transfer remains the most effective method to derive Embryonic Stem ( ES ) cells from adult tissues [74] , iSCNT using the oocytes of a more available species followed by the injection of these iSCNT-derived ES cells into a host blastocyst of the oocyte species could constitute an optimal route towards the generation of immuno-compatible organs of the donor species within the developing body of the recipient species [75] . A better understanding of the nucleocytoplasmic incompatibilities causing cybrid lethality may enable their correction and render such technology possible .
Xenopus laevis and Xenopus tropicalis adults were purchased from Nasco and maintained in our laboratory in separate systems , respectively operating at 18°C and 26°C . Eggs were collected dry onto dishes by gently massaging the frog's flanks [76] , since this was found to increase cross-fertilization efficiency . For nuclear inactivation , eggs were individually placed in a plastic dish with their animal pole facing up and submitted to UV irradiation for 25–30 s as previously described [32] . Fertilization or cross-fertilization was done by gently mixing the eggs with a crushed testis solution [76] . For sperm nuclear inactivation , a small clump-free volume taken from a crushed testis solution was spread on a glass slide to form a thin layer and exposed to UV irradiation for 20 s , using the same apparatus as for egg nuclear inactivation [32] . Such treatment resulted in 100% sperm DNA inactivation as determined by the characteristic haploid phenotype of the resulting embryos and examination of their nucleolar numbers . The optimal temperature range at which X . laevis and X . tropicalis are raised is different , but both species can develop normally at comparable rates at an intermediate temperature of 23°C [66] , [77] , and thus we have performed all of our experiments at this temperature . All embryos were maintained in 1/10 MMR unless otherwise mentioned , and staged according to the normal table of X . laevis development [78] . Cross-fertilized enucleated X . laevis eggs were de-jellied using a 2% L-Cysteine ( pH 8 ) solution , placed in a 6% Ficoll ( type 400 ) , 4/10 MMR solution [79] , and injected with distilled H2O ( dH2O ) , 460pg FRL-1 ( UTR- ) [58] , or 500 pg Xbra [80] in vitro synthesized capped mRNA in their animal half at the one-cell stage , or with dH2O or 2 . 3 pg recombinant human Activin A protein in their blastocoel at stage 8 , using a Drummond micro-injector . For karyotype analysis , embryos between stages 20 and 34 were cut open and immersed in dH2O for 20 min , fixed in 60% acetic acid for 5 min , and squashed onto a polysine microscope slide by pressing a coverslip down against it firmly . Slides were put on dry ice for 5 min , and then the coverslip was flicked off . The slide was then immersed in 20 µM Hoechst for 5 min , quickly drained , and 5 µl of vectashield mounting medium was added before covering with a new coverslip and sealing with nail polish . For immunostaining , embryos were harvested after 72 h of development and the dorsal ( yolk-free ) region was squashed on a polysine slide followed by freeze-cracking as above . They were then fixed in 4% paraformaldehyde for 30 min and blocked in 1% BSA for 1 h . Monoclonal mouse anti-fibrillarin ( AbCam; 1∶400 dilution ) and monoclonal Alexa 488-coupled anti-mouse ( Invitrogen; 1∶250 dilution ) secondary antibodies were used . DAPI ( 1 µg/ml ) was used as a counterstain . Eggs/embryos were de-jellied and collected at various time points after fertilization . Five eggs/embryos were transferred to an eppendorf tube , rinsed once in dH2O , and resupended in 250 µl of dH2O . Eggs/embryos were then homogenized by passing through a 26G syringe several times . The resulting suspension was chilled on ice and spun at 13 , 000 rpm for 5 min . 50 µl was then taken from the supernatant and transferred to a new tube where it was homogenized by pipetting . 25 µl of this was diluted 5-fold in dH2O , and kept on ice until assayed . ATP was then measured using an ATP Bioluminescent Assay kit ( Sigma ) ( at 1/500 dilution of the ATP assay mix ) according to the manufacturer's recommendation and a Glomax luminometer . Western blotting was performed essentially as previously described [38] . One X . laevis egg/embryo or 4 . 75 X . tropicalis eggs/embryos were loaded per well . ( This volume ratio was estimated based on the approximate diameter ratio ( 0 . 59 ) of the eggs of X . tropicalis/X . laevis utilized in our experiments . ) Rabbit monoclonal anti-phospho-AMPK ( Cell Signaling; 1∶1 , 000 dilution ) , mouse monoclonal anti-Ran ( BD Biosciences; 1∶2 , 000 dilution ) , rabbit polyclonal anti-Xbra ( raised against X . laevis Xbra ) [80] primary , and Alexa Fluor 680-conjugated goat anti-rabbit ( Invitrogen; 1∶20 , 000 dilution ) and 800CW-conjugated goat anti-mouse ( LI-COR; 1∶20 , 000 dilution ) secondary antibodies were used . Fluorescence was detected using the Odyssey detection system from LI-COR . Total RNA was prepared from single egg/embryo ( X . laevis ) or pools of three ( X . tropicalis ) using the RNeasy kit ( Qiagen ) and eluted in 20 µl . Total RNA concentration in each sample was determined by optical density at 260 nm . 10 µl of total RNA was used for real-time RT-PCR as previously described [81] using SYBR Green . Primers were as follows ( 5′→3′ ) : Vegt F: catcgctacaagcccaggtt , R: caatccccatggagaattgtaca , Xbra F: gaatgtgctggcaaagggtaa , R: ttccgttttcctgcatctttaaa , Chordin F: gctcagcaggtcacgcatgg , R: gttaggtatgtgcacttgtc , GATA4 F: gcttaaaactctcgccacaga , R: tgctttaagctaagaccaggttg , Mixer F: cagcagaggttcctgatgc , R: taagaggcaggaattccatggt . Vegt primers amplified both X . laevis and X . tropicalis sequences . Relative mRNA quantities were normalized for total RNA input for each sample before comparison . Explants were isolated in 1× MBS using fine forceps and a scalpel blade . Dorso-marginal zones were collected from stage 10 . 5 embryos essentially as previously described [82] . Animal-vegetal conjugates were constructed using stage 8–9 embryos as previously described [76] . For the activin conditioning assay , animal caps were collected from stage 8 embryos , incubated in 1× MBS with or without human recombinant activin A ( R&D Systems ) at various concentrations for various durations and washed . In every case , animal cap elongation was scored as previously described [50] , by the determination of their length∶width ratio after overnight incubation at 23°C in 1× MBS supplemented with antibiotics in 2% agarose-lined dishes , after verifying that sibling embryos had reached stage 19 .
|
When two species evolve separately for several million years , their respective genomes accumulate many small changes that together are responsible for the differences in their characters . Some of these affect the way eggs are prepared inside the germline , and/or how embryos develop , such that the egg cytoplasm of a given species can only support development promoted by its own genome or nucleus . Thus , developmental incompatibility arises between the cytoplasm and the nucleus of distant species during evolution and we don't know its mechanism . We have studied this phenomenon in an advantageous system using two evolutionarily distant frog species ( Xenopus laevis and Xenopus tropicalis ) . We found that hybrid frog embryos with X . laevis cytoplasm and X . tropicalis nuclei are always defective in an important process that is necessary to generate morphogenetic cell movements during development . Through a series of experiments in which we dissect out and/or recombine parts of such hybrid embryos and observe their behaviour in culture , we show that this phenomenon occurs because of malfunctions in the signalling cascade that is responsible for generating these cell movements . Thus , we postulate that inefficient molecular signalling contributes to the death of such hybrids .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology",
"embryology",
"evolutionary",
"biology",
"molecular",
"development",
"biology",
"morphogenesis",
"evolutionary",
"developmental",
"biology"
] |
2011
|
Deficient Induction Response in a Xenopus Nucleocytoplasmic Hybrid
|
The Asian tiger mosquito , Aedes albopictus , is a vector of dengue , Chikungunya , and Zika viruses . This mosquito inhabits a wide range of artificial water-holding containers in urban and suburban areas making it difficult to control . We tested the hypothesis that female-driven autodissemination of an insect growth regulator could penetrate cryptic oviposition habitats difficult to treat with conventional insecticidal sprays . Oviposition preferences of Ae . albopictus females for open and cryptic cups were tested in semi-field experiments . Two conventional larvicidal sprayers were tested to determine droplet penetration and larvicidal efficacy in open and cryptic habitats using Bacillus thuringiensis var . israelensis ( Bti ) in the field . Finally , the efficacy of pyriproxyfen autodissemination stations was assessed in cryptic and open cups in residential areas during 2013 and 2014 . Gravid females strongly preferred cryptic ( 53 . 1±12 . 9 eggs/cup ) over open ( 10 . 3±4 . 3 eggs/cup ) cups for oviposition . Cryptic cups showed limited droplet penetration and produced 0 . 1–0 . 3% larval mortality with a conventional backpack and low-volume sprays of Bti . The autodissemination stations effectively contaminated these cryptic cups ( 59 . 3–84 . 6% ) and produced 29 . 7–40 . 8% pupal mortality during 2013–2014 . Significant pupal mortality was also observed in open cups . The autodissemination station effectively exploits the oviposition behavior of wild gravid females to deliver pyriproxyfen to targeted oviposition habitats . Although the pupal mortality in cryptic cups was relatively lower than expected for the effective vector control . Autodissemination approach may be a suitable supporting tool to manage Ae . albopictus immatures in the cryptic habitats those are less accessible to conventional larvicidal sprays .
Aedes albopictus is a highly invasive mosquito species that has originated from South East Asia and now has spread in most part of the world[1] . In the United States , this mosquito species was first reported from Texas in 1985 and has since expanded its range to more than 30 states [2] . Aedes albopictus ( Skuse ) is a major public health problem due to its ability to transmit the dengue , chikungunya , yellow fever viruses , and is a competent vector for at least 22 other arboviruses including Zika [3] . There are more than 50 million dengue cases annually , and around 2 . 5 billion peoples are living at the risk of infection [4] . Source reduction , adulticides and larvicide applications are routinely used to manage Asian tiger mosquitoes [5] . Adulticides are very effective in suppressing adult populations but populations tend to quickly rebound [6 , 7] . Source reduction by eliminating potential larval habitats is also a key element of mosquito control . Eliminating the diverse array of containers used by peridomestic Ae . albopictus , including discarded tires , trash , soda cans , plant pots , bird baths , corrugated drain pipes etc . , is extremely challenging and labor intensive [5 , 6] . In addition , mosquito control agencies find it difficult to get access to the private properties where these habitats are located [8] . Moreover , accumulation of containers such as beverage cans and other trash are hard to manage because they are often rapidly re-generated . Larvicide application is an effective alternative to control immature Ae . albopictus in larval habitats . Larvicides are applied with backpack or larger motorized sprayer such as a truck-mounted low and ultralow volume sprayers [9 , 10] . For instance , Bacillus thuringiensis var . israelensis is a well-established larvicide used for mosquito control , particularly in urban areas , with proven efficacy [8 , 10] . It is imperative for any larvicide application to reach concealed cryptic habitats , where this mosquito prefers to oviposit . However , the effectiveness of conventional larvicide equipment to deliver larvicide in such cryptic habitats is not well documented . There is no information available regarding the efficacy of conventional methods to deliver larvicides into cryptic larval habitats . Autodissemination using the insect growth regulator pyriproxyfen , a juvenile hormone mimic , has recently emerged as a potential new vector management tool . This “pull-and-push” method exploits the skip oviposition behavior [11] of the Asian tiger mosquito to treat water-holding containers [12] . We have developed an autodissemination station that attracts gravid females and contaminates them topically with pyriproxyfen as they exit the station to locate new oviposition sites where the insecticide is then transferred [12] . We ask if Ae . albopictus prefers cryptic oviposition cups over the open cups , and that if cryptic oviposition cups are better targeted by autodissemination approach compared to conventional spray technologies . We examined Ae . albopictus oviposition preference for cryptic and open cups in a large cage field experiment; tested the efficacy of larvicidal sprayers to treat cryptic cups through droplet penetration and larvicide application , and then examined autodissemination efficacy in open versus cryptic cups against field populations of the Asian tiger mosquito .
A laboratory colony of Ae . albopictus mosquito was maintained at the Center for Vector Biology , Rutgers University . Eggs were originally field collected from Mercer and Monmouth Counties , New Jersey in 2008 . Mosquitoes were reared at 26 ± 1°C temperature and 75 ± 5% relative humidity under 16L:8D photoperiod . Brewer’s yeast was provided as a larval food ( 30 mg/L ) daily . Adult mosquitoes were fed on 10% sucrose solution and guinea pigs were used for the blood feeding . Previously , a baseline data for the susceptibility of the laboratory-raised colony of Ae . albopictus to pyriproxyfen has been established [13] . The study protocol was approved by the Institutional Animal Care and Use Committee ( IACUC ) , Rutgers University . The study was performed in strict accordance with IACUC Animal Use Protocol # 86–129 . The oviposition preferences of gravid Ae . albopictus females were tested in a semi-field experiment conducted in a 50 ( L ) x 3 ( W ) x 2 ( H ) m mosquito proof tunnel cage ( Fig 1A ) in an open area surrounded by trees and bushes . Five plastic cups ( 450 ml capacity ) were used as open oviposition cups ( Fig 1B ) , whereas another five cups inserted into plastic pipes ( 45 ( L ) x 15 ( diam ) cm ) were used as cryptic cups ( Fig 1C ) . Each sentinel cup held 250 ml dechlorinated tap water and was lined with filter paper ( Whatman filter paper #1 ) as an oviposition substrate . Open and cryptic cups were placed alternatively every 10 m distance at 15 cm away from the cage side walls ( Fig 1A ) . Two hundred gravid females ( 4–5 days post-blood feeding ) were released at the center of the cage . After 5 days , the cups were transported to the lab . Oviposition preference was determined by counting the number of cups that received eggs and the number of eggs that were accumulated in each cup . To nullify positional preference , open and cryptic cups were switched in each experiment . The experiment was repeated three times during June–July 2015 . Two conventional larvicide applicators , a Stihl® SR 420 backpack sprayer ( Andreas Stihl Ag & Co . KG , Waiblingen , Germany ) and a truck-mounted Curtis Dyna-Fog Ag-Mister LV- 8™ ( Curtis Dyna-Fog , Westfield , IN , USA ) low-volume ( LV ) sprayer were evaluated for their efficacy to deliver spray droplets in open and cryptic cups . Tests were carried out in an open parking lot at Rutgers University . Cryptic cups ( i . e . , ovicups inserted into pipes ) were placed in parallel , perpendicular and oblique positions relative to the equipment’s spray direction . Cups were placed 1 m distance apart in three replicates . Kromekote papers ( Mohawk Fine Papers , Inc . , Cohoes , NY , USA ) measuring 5×7 . 5 cm were placed in the bottom of each cup . A 2% solution of Red 40 granular dye ( Glanbia Nutritionals , Carlsbad , CA , USA ) was sprayed using the backpack and truck-mounted sprayers . Experiments were conducted on the day when the prevalent wind speed was minimal and was found to be between 2 to 5 km/h . Backpack sprayer was calibrated to flow rate of 0 . 12 L/min and application was performed from the distance of 3 m from the cup placement . The low-volume sprayer was adjusted to flow rate of 8 . 3 L/min and application was performed at 10 m from the cup placement . Both sprayers moved at the speed of 8 km/h and perpendicular to the wind direction . Papers were collected after 30 min of application and analyzed by DropVision AG system ( Leading Edge Associates , Inc . , Waynesville , NC ) to determine droplet size and density . Each experiment was repeated three times . Two field trials were carried out to evaluate Bacillus thuringiensis var . israelensis ( Bti ) ( VectoBac® WDG , Valent BioScience Corp . , Libertyville , IL , USA ) larvicide penetration in open and cryptic cups in Essex County ( New Jersey , USA ) in 2013 with a backpack sprayer and in Mercer County ( New Jersey , USA ) in 2014 with truck-mounted LV sprayer . Whereas the Essex trial was conducted in a suburban area , the Mercer trial was performed in an inner city urban area . The study areas were divided into one control and three treatment plots of ca . 4000 m2 per plot . Control plot was 1000 m apart from the treatment plots . Open and cryptic cups ( 12 cups per plot ) were placed in shaded areas and filled with 250 ml of dechlorinated water . Larvicide applications were then made by the local county mosquito control agencies under the authority of Title 26 . 9 of the New Jersey administrative code . Larvicide was applied at the application dose of 1 kg/ha with flow rate of 0 . 12 L/min and 8 . 3 L/min for backpack and LV sprayer , respectively . Both the sprayers were moving at the speed of 8 km/h . The control plot received five open cups . Cups were collected 30 min post-application and transported to the lab for larval bioassays with each cup receiving 20 early 3rd instar Ae . albopictus larvae . A separate laboratory control group was tested concurrently in three replicates in 250 ml dechlorinated water . Larval mortality was recorded after 24 h with moribund larvae recorded as dead . Larval food ( Brewer’s yeast , 30 mg/L ) was provided during the bioassay . Autodissemination studies were conducted on the same Mercer and Essex Counties experimental plots as the sprayer experiments . Each 4000 m2 treatment plot was divided into four ca . 1000 m2 test blocks , with each block receiving one station in 2013 and two stations in the 2014 study . Dual phase autodissemination stations [14] were used in 2013 , whereas the stations tested in 2014 were slightly modified . The major differences between 2013 and 2014 stations were the configuration of formulation cartridges . The autodissemination station in 2013 contained a circular dual band formulation cartridge ( inner oil and outer powder bands ) situated at the top of the transfer chamber ( Fig 2A and 2B ) . To reduce insecticidal load per autodissemination station and economical cost , a dual layer formulations cartridge situated at the side wall of transfer chamber was used in 2014 ( Fig 2C and 2D ) . Instead of single formulation ring in 2013 , dual layer cartridge system in 2014 provided an opportunity for the station to perform even if one formulation layer fails under field conditions . Autodissemination stations were loaded with oil ( 20% a . i . ) and powder ( 60% a . i . ) pyriproxyfen formulation . Open and cryptic cups ( 12 each ) were placed in the treatment plot under foliage but away from direct sunlight . The control plot received five open sentinel cups . To increase the oviposition attractancy , stations and sentinel cups were filled with 1000 ml and 250 ml of oak infusion , respectively [15] . The day stations were deployed was considered “week 0” . Water samples were collected at weeks 2 , 4 and 8 in 2013 , whereas , an additional sample was collected at week 12 in 2014 . At each collection event , infusion from the field ovicups was transferred to a new container after swirling , and the cups were refilled with 250 ml of infusion and returned to the field . Cups with less than 250 ml of infusion due to evaporation were replenished with tap water to the required volume before sampling . Collected samples were transported to the laboratory and tested for pyriproxyfen activity by larval bioassay . Samples were filtered in the lab with a paper towel ( Bounty® Procter and Gamble , USA ) to remove organic matter and wild mosquito populations . From each sample , 200 ml of infusion was taken for bioassay to determine insecticidal activity and the remaining 50 ml was stored in an amber colored glass bottle at -20°C for subsequent residue analysis . Each bioassay container received 20 early 3rd instar Ae . albopictus larvae . Pupal mortality was recorded as pyriproxyfen activity and larval mortality was excluded from analyses . Incomplete emergence or dead adult with attached exuviae were also considered as pupal mortality . For each experiment , an additional laboratory control was placed in three cups filled with dechlorinated water . Brewer’s yeast ( 30 mg/L ) was provided as larval food twice a week . Cups were observed until either all individuals emerged as adults or died . Samples showing complete emergence inhibition or 100% pupal mortality were sent for residue analysis at Golden Pacific Laboratory , California , USA using LC-MS-MS analysis as described previously [14] . Data were first tested for normal distribution using Shapiro-Wilk test . Data are presented as mean ± SE unless otherwise stated . Mann-Whitney Rank sum test was performed to compare eggs received by open and cryptic cups in tunnel experiment . Fisher Exact test was used to compare oviposition receiving and non-receiving , cryptic and open cups . Mann-Whitney Rank sum tests were performed for pairwise analysis to examine statistically significant differences for droplet density and median droplet size between open and cryptic cups . Efficacy of larvicide and pyriproxyfen autodissemination was expressed as percent larval and pupal mortality , respectively . In all bioassay experiments , control mortality was adjusted with open and cryptic cups mortality using Abbott’s formula [16] . Mann-Whitney tests were used to determine the significant difference in larval mortality between open and cryptic cups during larvicide application , and in pupal mortality during autodissemination experiments . Percent pyriproxyfen contamination was calculated using the following formula: %pyriproxyfencontamination=NumberofcontainersshowingpyriproxyfenactivityTotalnumberofcontainers×100
Aedes albopictus females showed a strong preference for oviposition in cryptic cups compared to open cups . Females deposited 83 . 8% ( 53 . 1±12 . 9 eggs/cup ) of their eggs in cryptic cups , whereas open cups received only 16 . 2% ( 10 . 3±4 . 3 eggs/cup ) ( Mann-Whitney , T = 160 , p = 0 . 002 ) . Nearly 93% of cryptic cups received eggs , which was significantly higher than the open cups ( 40% ) ( Fisher Exact , p = 0 . 005 ) . Backpack and truck mounted sprayers were highly effective at delivering spray droplets to open cups but ineffective at delivery to cryptic cups ( Fig 3 ) . With the backpack sprayer , open cups received the highest droplet density ( 40 . 3±7 . 9 droplets/cm2 ) followed by cryptic cups hidden inside pipes in perpendicular ( 11 . 7±7 . 1 droplets/cm2 ) , oblique ( 11 . 1±3 . 3 droplets/cm2 ) and parallel ( 3 . 0±2 . 5 droplets/cm2 ) positions ( Fig 3A ) . Irrespective of their position relative to spray direction , droplet density was significantly lower in cryptic than open cups ( Mann-Whitney , parallel , T = 55 , p = 0 . 001; oblique , T = 80 , p = 0 . 019; perpendicular , T = 74 , p = 0 . 008 ) . Within cryptic cups , there was no significant difference between parallel and perpendicular cryptic cups and between oblique and perpendicular cups , however , droplet density in parallel cups was significantly lower than the oblique cups ( Mann-Whitney , T = 58 , p = 0 . 017 ) . Median droplet diameter from the backpack sprayer was 174 . 82 μm ( range , 51 . 37 to 390 . 66 μm ) ( Fig 3A ) . Droplets received by cryptic cups inside parallel and oblique pipes were smaller than the droplets received by open cups ( Mann-Whitney , parallel , T = 68 , p = 0 . 003; oblique , T = 74 , p = 0 . 008 ) . Droplet sizes received by open and cryptic cups in the perpendicular pipes did not differ significantly . Similar droplet penetrations were observed from the truck mounted LV sprayer ( Fig 3B ) . Open cups received the highest droplet density ( 2 . 72±0 . 9 droplets/cm2 ) followed by cryptic cups in perpendicular ( 0 . 15±0 . 03 droplets/cm2 ) , parallel ( 0 . 11±0 . 03 droplets/cm2 ) and oblique ( 0 . 06±0 . 01 droplets/cm2 ) pipes . Irrespective of position , droplet density in cryptic cups was significantly lower than open cups ( Mann-Whitney , parallel , T = 45 , p≤0 . 001; perpendicular , T = 46 , p≤0 . 001; oblique , T = 45 , p≤0 . 001 ) . There was no significant difference in droplet density within cryptic cups regardless of their position . Median droplet size of droplets produced by the LV sprayer was 95 . 0 μm diam ( range , 12 . 67 to 180 . 75 μm ) ( Fig 3B ) . The size of droplets received by cryptic cups in the oblique pipe was significantly smaller than the open cups ( Mann-Whitney , T = 45 , p≤0 . 001 ) . The backpack application of B . thuringiensis larvicide produced 75 . 8±7 . 0% Ae . albopictus mortality in open ovicups , whereas cryptic cups showed virtually no measurable mortality at 0 . 3±0 . 3 ( Mann-Whitney , T = 1867 , p≤0 . 001 ) . The truck-mounted treatment was similarly ineffective against cryptic cups ( 0 . 1±0 . 1% mortality ) in comparison to open cups ( 19 . 4±6 . 1% ) ( Mann-Whitney , T = 1499 , p = 0 . 001 ) . Cups placed in the control plot did not show larval mortality in these trials . In the 2013 autodissemination experiment , mean percent pupal mortality in open cups was 8 . 4% ( range , 0 . 9 to 29 . 6% ) with highest pupal mortality was recorded in week 8 ( 29 . 6% ) samples ( Fig 4A ) . In cryptic cups , mean percent pupal mortality was 13 . 4% ( range , 0 . 4 to 40 . 8% ) with highest pupal mortality in week 8 samples ( Fig 4A ) . No significant difference in pupal mortality was found between open and cryptic cups in weeks 2 and 4 , although week 8 samples showed significantly greater mortality in cryptic than open cups ( Mann-Whitney , T = 136 , p = 0 . 03 ) ( Fig 4A ) . Mean pyriproxyfen contamination in cryptic and open cups was 59 . 3±4 . 6% ( range , 8 . 3 to 100% ) and 46 . 0±8 . 2% ( range , 8 . 3 to 75% ) respectively , however , the difference was not significant ( Fig 4B ) . Autodissemination was further confirmed by residue analysis from samples exhibiting high pupal mortality , with twice the concentration found in cryptic as compared to open cups ( 1 . 64 and 0 . 87 μg/L respectively ) . The number of autodissemination stations in treatment plots was increased in 2014 to increase the probability of pyriproxyfen transfer . Mean percent pupal in open cups was 13 . 4% ( range , 7 . 5 to 26 . 6% ) , whereas in cryptic cups it was 29 . 7% ( range , 11 . 0 to 41 . 8% ) ( Fig 5A ) . Pupal mortality in sentinel cryptic cups was consistently greater than open cups throughout the study ( Fig 5A ) ( Mann-Whitney , week 2 , T = 711 , p≤0 . 001; week 4 , T = 942 , p = 0 . 025; week 8 , T = 486 , p≤0 . 001; week 12 , T = 583 , p = 0 . 04; overall , T = 15543 , p≤0 . 001 ) . Pyriproxyfen contamination ranged from 36 . 1 to 88 . 4% ( mean , 58 . 2±6 . 9% ) in open cups and 72 . 9 to 96 . 7% ( mean , 84 . 6±5 . 3% ) in cryptic cups ( Fig 5B ) . Residue analysis confirmed the transfer of pyriproxyfen from contaminated mosquitoes into open and cryptic cups at 0 . 0046 and 0 . 0103 μg/L , respectively .
Asian tiger mosquitoes can use almost any small , water-holding container for larval development; however , this species showed a strong oviposition preference for cryptic over open cups in our field cage experiment suggesting a higher contribution of cryptic habitats in population buildup . The major difference between open and cryptic cups is their accessibility , open cups have a highly apparent and accessible top entrance , whereas cryptic cups are hidden within a pipe and only accessible from side entrances . Greater Ae . albopictus larval prevalence in corrugated extension tubes than open ( i . e . , exposed ) containers in urban residential areas has been previously reported [17] . Higher larval prevalence was found in containers such as tires , water bottles , tin cans , gutter hoses , corrugated pipes , and water drainage pipes etc . [6 , 18–22] which displayed some degree of a cryptic nature . If our hypothesis is correct that cryptic habitats hold much or most of the Ae . albopictus population , effective control strategies to curtail container mosquitoes in these habitats would be vital . Yet the droplet data demonstrated for the first time that cryptic cups are nearly impenetrable to the conventional backpack and truck mounted applications routinely used for larvicide applications [9 , 23] . Irrespective of their position relative to the spray application path , cryptic cups were difficult to reach by aqueous droplets . Similarly , cryptic cups were nearly invulnerable to our Bti applications as reflected in almost negligible larval mortality compared to open cups . The side entrance of cryptic cups makes them almost inaccessible to insecticidal droplets suggesting that the efficacy of conventional sprayers against container mosquitoes may be limited and their efficacy is overestimated when evaluated against open cups only as the majority of the population preferred cryptic habitats . For effective management of the Asian tiger mosquito , an alternate vector management strategy is needed which can specifically target cryptic habitats . Autodissemination is emerging as a potential new strategy with a promise to deliver insecticides to both exposed and hidden habitats . Autodissemination exploits the behavior of the insect as a vehicle to deliver toxicants to the pest habitats . Consider the goat moth , Coccus coccus , which also lives in cryptic , hidden habitats . Pre-infecting adults with entomopathogenic nematodes and then releasing them to locate and deliver the parasites to these hard-to-reach target sites resulted in 86% larval mortality , whereas conventional methods produced only 4% larval mortality [24] . Similarly , the ability of autodissemination to transfer lethal concentrations of insecticides has been successfully demonstrated in the laboratory and the field for Ae . albopictus and other mosquito vectors [12 , 25–29] . Previous studies , however , have used only open cups to assess autodissemination efficacy and did not examine efficacy in cryptic habitats . We have tested and demonstrated that autodissemination can deliver toxic concentrations of insect growth regulator to both open and cryptic cups . We have developed multiple designs for autodissemination stations that exploit the oviposition behavior of gravid Ae . albopictus to deliver pyriproxyfen to larval habitats [12 , 14 , 30] . Attracted to the station by an oak infusion , the mosquito enters the station but is unable to reach the water and oviposit . In the current design , when they exit they are forced to walk across an oil and powder fabric bands impregnated with pyriproxyfen . In addition to its insecticidal role , the oil formulation serves additional roles as a sticker to enhance attachment of the powder formulation , and to facilitate quick release into the larval habitat during oviposition . In two years of field studies , we have shown here that autodissemination stations are useful against both open and cryptic cups . We observed higher pupal mortality and pyriproxyfen contamination in cryptic than open cups , however , this difference was not significant in 2013 . Although pyriproxyfen reaches both cryptic and open cups in 2013 , overall field efficacy was at best modest that year . Our 2013 plots were located in a suburban area with unmistakably low mosquito populations; because autodissemination relies entirely on wild adult mosquitoes to transfer the insecticide , the outcome from low populations cause low mortality . Greater autodissemination efficacy was obtained in the 2014 field study because stations were modified to perform for a longer period , the number of stations per block was doubled , and the trial was conducted in an inner-city area where adult populations were chronically high . Pyriproxyfen autodissemination in cryptic cups was consistently higher than open cups during the entire study period . Mortality in 2014 was comparable to previous autodissemination field experiments [30] , though the pupal mortality in cryptic cups was relatively low to reach the desired level of effective control recommended by vector control authorities . Efficacy may have been underestimated because field samples were filtered to remove organic debris which likely reduced pyriproxyfen given this chemical’s propensity for adsorption to different substrates [31] . Pyriproxyfen dissemination was further confirmed by residue analysis of the sentinel cups , while previous autodissemination field studies have tended to rely on larval bioassay to show pyriproxyfen presence and activity but lacked direct evidence of transfer to sentinel cups [25 , 27 , 28] . Although , open and cryptic cups have shown the presence of pyriproxyfen , we found a surprisingly large difference in pyriproxyfen concentration between the 2013 and 2014 samples . One possible explanation for such a huge difference may be variation in the samples . During 2013 , we analyzed week 8 samples for residue analysis . Whereas in 2014 , week 4 samples were analyzed instead due to low pupal mortality in week 8 and week 12 samples . There is a possibility that pyriproxyfen may have accumulated in field sentinel cups over time in 2013 , which eventually resulted in higher concentrations . However , this requires further investigation to establish the relationship of various factors i . e . age of water sample , organic components , temperature , and mosquito density that can affect pyriproxyfen residual concentrations . Autodissemination system is a low maintenance and easy to operate which can cover an entire mosquito season once deployed . Pyriproxyfen is a highly effective juvenile hormone analog , besides its pupicidal action , pyriproxyfen sterilizes adult females and decreases spermatogenesis in males [32] , shows ovicidal activity [13] , and can terminate egg diapause prematurely [33] . The impact of pyriproxyfen on non-targets is minimal as dissemination is tightly targeted to small , scattered , container habitats . Additionally , the amount carried by females is minuscule which further reduces insecticide load in the environment [14] .
Conventional methods of evaluating larvicidal sprays through the placement of open sentinel cups may overestimate efficacy if much of the container larval population are found in their preferred cryptic habitats . Greater consideration to cryptic habitats may improve mosquito control programs and likely to have limited efficacy without specifically targeting cryptic habitats . Our results provide evidence that female-driven autodissemination can be a viable control strategy to suppress Ae . albopictus mosquitos in hidden habitats which are otherwise difficult to reach by conventional methods . Although autodissemination is not likely an area-wide approach , it shows great prospects for treating ‘hot spots , ’ particularly where larval habitats are difficult to locate and where mosquito control agencies lack access due to restricted entry . Autodissemination stations provide an opportunity to enhance larval site treatment coverage for mosquito control and may be a powerful tool to enhance integrated vector management and target-based insecticide applications .
|
Aedes albopictus is a highly invasive mosquito species and a competent vector of dengue , Chikungunya , Zika and other arboviruses . In the absence of therapeutics , vector control is the only means of controlling these diseases . Larvicides are important tool to limit mosquito populations . Backpack and low volume sprayers are heavily used for ground-based larvicide applications , but their potential to reach hidden or cryptic larval habitats is not well established . In the present study , we found that Ae . albopictus shows a strong oviposition preference for cryptic cups . Moreover , cryptic cups received limited droplet penetration using conventional spray methodologies , resulting in almost negligible larval mortality as compared to open cups in field experiments . We further evaluated the efficacy of an autodissemination station to deliver the insect growth regulator pyriproxyfen to cryptic cups in field experiments . Autodissemination shows good efficacy in transferring toxic concentrations of pyriproxyfen to cryptic cups and may provide a significant advantage over sprayers in delivering insecticide to the cryptic larval habitats strongly preferred by Ae . albopictus .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Conclusions"
] |
[
"larvicides",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"togaviruses",
"pathogens",
"microbiology",
"animals",
"alphaviruses",
"reproductive",
"physiology",
"viruses",
"chikungunya",
"virus",
"rna",
"viruses",
"oils",
"pest",
"control",
"insect",
"vectors",
"agrochemicals",
"lipids",
"geography",
"medical",
"microbiology",
"epidemiology",
"microbial",
"pathogens",
"disease",
"vectors",
"insects",
"agriculture",
"arthropoda",
"pesticides",
"insecticides",
"biochemistry",
"mosquitoes",
"urban",
"areas",
"physiology",
"viral",
"pathogens",
"oviposition",
"earth",
"sciences",
"biology",
"and",
"life",
"sciences",
"geographic",
"areas",
"organisms"
] |
2016
|
Targeting a Hidden Enemy: Pyriproxyfen Autodissemination Strategy for the Control of the Container Mosquito Aedes albopictus in Cryptic Habitats
|
In budding yeast , activation of many DNA replication origins is regulated by their chromatin environment , whereas others fire in early S phase regardless of their chromosomal location . Several location-independent origins contain at least two divergently oriented binding sites for Forkhead ( Fkh ) transcription factors in close proximity to their ARS consensus sequence . To explore whether recruitment of Forkhead proteins to replication origins is dependent on the spatial arrangement of Fkh1/2 binding sites , we changed the spacing and orientation of the sites in early replication origins ARS305 and ARS607 . We followed recruitment of the Fkh1 protein to origins by chromatin immunoprecipitation and tested the ability of these origins to fire in early S phase . Our results demonstrate that precise spatial and directional arrangement of Fkh1/2 sites is crucial for efficient binding of the Fkh1 protein and for early firing of the origins . We also show that recruitment of Fkh1 to the origins depends on formation of the pre-replicative complex ( pre-RC ) and loading of the Mcm2-7 helicase , indicating that the origins are regulated by cooperative action of Fkh1 and the pre-RC . These results reveal that DNA binding of Forkhead factors does not depend merely on the presence of its binding sites but on their precise arrangement and is strongly influenced by other protein complexes in the vicinity .
Replication of genomic DNA in budding yeast ( Saccharomyces cerevisiae ) is initiated from hundreds of origins throughout the S phase . Replication origins can be characterized by their efficiency , which refers to the probability that a particular origin will fire in a given cell cycle , and by the timing of their firing in the S phase . In general , early firing origins are also efficient , i . e . replication is initiated from these origins in almost every S phase . However , the determinants of early origin firing are not fully understood . It has been shown that origins located in euchromatic regions close to centromeres fire early in the S phase , while origins found in sub-telomeric heterochromatin are generally late-firing [1 , 2] . Relocation of several origins into ectopic loci has revealed that some origins adjust their firing time according to the local chromatin context , while a set of chromosomal localisation-independent origins retain their early-firing pattern in the new location [3] . Both origin relocations and genome-wide DNA replication initiation studies have shown that Forkhead transcription factor family members Fkh1 and Fkh2 are required to ensure early firing of chromosomal localisation-independent origins . These findings were further confirmed by the fact that these origins contain at least two consensus binding sites for Fkh1/2 proteins , as well as by observations that disruption of these sites leads to a loss of the origin’s early firing signature [3 , 4] . The consensus binding sequence for Forkhead family proteins , RYMAAYA , is rather loosely defined and allows many variations in the actual DNA sequence [5] . Therefore , approximately 46 , 000 Fkh1/2 consensus sequences can be found throughout the budding yeast genome , but only about 1650 of them are actually bound by Forkhead factors [6] . Remarkably , overexpression of Fkh1 leads to its recruitment to multiple new loci that were not occupied at the normal expression level of Fkh1 [7] , indicating that availability of free Forkhead proteins might be limiting in cells . This also suggests that in the presence of great excess of potential binding sites the Forkheads are stably recruited only to the loci that either are more accessible , or where their binding is supported by other proteins . It is not clear how the binding of Fkh proteins is regulated in replication origins . However , mutation of one of the two Fkh sites present at ARS305 , ARS607 and ARS737 significantly reduces binding of Fkh1 to these origins [3] . This suggests that multiple Fkh1/2 consensus sites in close proximity to each other are required for efficient binding of Fkh1 . In addition , due to the asymmetrical nature of the Fkh1/2 consensus binding sequence , the efficient binding of Forkhead factors may be influenced by orientation of the sites . Interestingly , three chromatin-independent early-firing replication origins ARS305 , ARS607 and ARS737 , contain two Fkh1/2 sites in divergent orientation relative to each other and separated by 72 base pairs , suggesting that the overall configuration of the Forkhead binding motifs at early replication origins is carefully conserved . To elucidate the role of the precise orientation and spacing of Fkh1/2 sites in regulation of early-firing replication origins , we tested the efficiency of Fkh1 binding to origins with altered patterns of Fkh1/2 consensus sites , as well as the firing profile of such origins . We show that only wild-type configuration of Forkhead binding sites leads to efficient Fkh1 recruitment to an origin in G1 and to early firing of the origin in the S phase . We also demonstrate that even when the consensus sites are present in their wild-type configuration , Fkh1 fails bind to the origin if recruitment of the Mcm2-7 complex is disrupted . These results suggest that efficient DNA binding of Forkhead family proteins is strongly influenced by cooperative interactions with other DNA binding factors and is not determined merely by their consensus DNA binding sequence .
Our previous study has revealed that two Fkh1/2 binding sites were required for early firing of replication origins ARS305 , ARS607 and ARS737 [3] . At all these origins , one of the Fkh1/2 binding sites is located close to the ARS consensus sequence ( ACS ) and another is found 72 base pairs away . To test whether the distance between the Fkh1/2 sites is important for efficient binding of Fkh1 and for origin regulation , we made a panel of yeast strains with modified ARS607 origins inserted into a GAL-VPS13 locus . In all modified ARS607 sequences , the native ACS-proximal Fkh1/2 site ( partially overlapping with ACS ) was left undisturbed , while the distal Fkh1/2 site was mutated ( Fig 1A ) . Mutation of the distal consensus sequence led to a significant drop in Fkh1 occupancy , although it remained higher than at loci that do not contain any Fkh1/2 binding sites ( Fig 1C ) . We then introduced a new Fkh1/2 site at various distances from the proximal site and determined the efficiency of Fkh1 binding to these origins in G1-arrested cells . When the Fkh1/2 sites were separated by 10 base pairs , binding of Fkh1 protein to the origin was detected at slightly reduced levels compared to the wild-type ( wt ) sequence ( Fig 1C ) . In contrast , binding of Fkh1 to all other origins was indistinguishable from that seen at an origin where only a single Fkh1/2 site was present , suggesting that precise distance between Fkh1/2 binding sites in ARS607 is critical for efficient recruitment of Fkh1 protein to the origin ( Fig 1C ) . To finely probe the tolerance of Fkh binding to altered size of the gap between Fkh1/2 sites , we made 5 bp , 10 bp and 15 bp insertions and two different 10 bp deletions between the Fkh1/2 binding sites in GAL-VPS13-ARS607 ( Fig 1B and S4 Fig ) . In these constructs , modifications of ARS607 were minimal and importantly , all modified loci retained the original Fkh1/2 sites in their immediate surrounding sequences . This minimized the possibility that recruitment of Fkh1 was affected by local DNA sequence context rather than by the change in gap size between the Fkh1/2 sites . We found that all introduced modifications , even insertion of 5 bp into the locus , caused significant drop of Fkh1 binding to the origin ( Fig 1C ) . Concordant with their impaired ability to bind the Fkh1 protein , origins ARS305 , ARS607 and ARS737 with mutated Fkh1/2 binding sites also lose their early firing pattern in vivo [3] . To determine whether the distance of Fkh1/2 sites in ARS607 is crucial also for early firing of the origin , we arrested cells in G1 and released them synchronously into S phase in the presence of hydroxyurea ( HU ) , thus enabling the firing of early but not late origins . The dynamics of the relative copy number of the GAL-VPS13-ARS607 locus following the release from G1 block were determined by qPCR analysis of extracted genomic DNA . If the origin could fire early in S phase , the locus initiated replication in the presence of HU and the relative amount of its DNA was expected to increase during the experiment . This assay revealed that only the strain with wt ARS607 was able to support early replication of the GAL-VPS13-ARS607 , while in all other strains the locus was not replicated in the presence of HU ( Fig 1D and S1 Fig ) . Interestingly , the origin where the two Fkh1/2 sites were separated by 10 bp did not support early replication of the locus , despite the fact that Fkh1 occupancy at that origin was relatively high ( Fig 1C ) . This suggests that the proper spacing of Fkh1/2 binding sites rather than the mere binding of Fkh1 is critical for the early firing of the origin . To show that the alterations in ARS607 sequence that affected its early firing did not render it inactive , we confirmed that the Mcm2-7 complex was recruited to all modified GAL-VPS13-ARS607 loci , indicating their proper licensing ( Fig 1E ) . In early firing origins ARS305 , ARS607 , and ARS737 the asymmetrical Fkh1/2 consensus sequences are found in divergent orientation , suggesting that directionality , or symmetry of Fkh1/2 sites may be important for efficient binding of Forkhead factors to replication origins . To test this hypothesis , we reversed the orientations of Fkh1/2 consensus sites in ARS305 ( Fig 2A ) and tested whether this affected the efficiency of Fkh1 binding and early firing of the origin . Fkh1 binding was nearly lost in all Fkh1/2 binding site reversal mutants , including the one in which both sites were rotated to form a convergent conformation , implying that the proper orientation of Fkh1/2 sites is critical for Fkh1 binding to ARS305 ( Fig 2B ) . As expected , none of the above mutants were able to fire early in S phase when cells were released into HU-containing media ( Fig 2C and S2 Fig ) , although all of them were properly licensed ( Fig 2D ) . As the Forkhead consensus sequence RYMAAYA allows numerous variations in the actual DNA sequence , these sites can be found frequently throughout the genome . Our results with modified origins ARS607 and ARS305 suggest that the two Fkh1/2 sites must be present in proper orientation and with correct spacing between them for Forkhead-dependent regulation of these origins . To find out how many double Fkh1/2 sites are present in the yeast genome and how many of them co-localize with early replication origins , we searched the yeast genome for locations of Fkh1/2 consensus binding sites and plotted those onto a genome-wide early DNA replication initiation profile , based on BrdU incorporation into DNA in the presence of HU [8] . Specifically , we sought loci where two Fkh1/2 consensus sites were separated by 62–88 base pairs and oriented in three different patterns–divergently , convergently , or unidirectionally . After analyzing the entire yeast genome , we found 5023 double Fkh1/2 sites that were separated by 62 to 88 bp and oriented in one of the three possible patterns ( S1 Table ) . Approximately 3% of these patterns co-localized with early replication origins . However , when both the distance and the orientation of sites were taken into account , the sites in divergent orientation separated by 71–79 bp were almost three times more likely to co-localize with early origins than sites with alternate configurations ( Fig 3A ) . When similar analysis was performed using late replication origins , no overrepresentation of any pattern of Fkh1/2 sites was found . Moreover , divergently oriented sites separated by 71–79 bp were visibly underrepresented in late origins ( Fig 3B ) . Next , we mapped the locations of ARS consensus sequences in early origins that overlapped with divergent Fkh1/2 sites separated by 71–79 bp . The ACS was found in close proximity ( up to 100 bp ) to Forkhead binding sites at 20 origins , and alignment of these revealed several common features ( Fig 3C and S3 Fig ) : First , there was no clear preference for position of the ACS within the Fkh motif–it could be found between or outside of the two sites; however , very often ACS partially overlapped with one of the Forkhead consensus sequences . Secondly , the ACS and its proximal Fkh1/2 consensus site were located on complementary strands , i . e . ARS consensus on the T-rich strand and the proximal Forkhead site on the A-rich strand . Thirdly , continuous adenine nucleotide tracks were present between the Fkh1/2 sites . All analysed origins were found to contain either at least one continuous A-track of at least 5 bases , or multiple 4-bp A-tracks . We speculate that these sequences facilitate initial melting of DNA strands during origin activation . As Fkh1/2 sites are located very close to the ACS in early replicating origins , we tested whether the formation of the pre-RC influences the efficiency of Fkh1 recruitment to these loci . We mutated the ACS sites in VPS13-ARS305 and VPS13-ARS607 , as well as in native ARS305 and ARS737 loci ( Fig 4A ) , and measured Fkh1 binding to these origins . Although the sequence , spacing and orientation of Fkh1/2 sites were unchanged , the binding of Fkh1 was nearly lost in all ACS-mutated loci ( Fig 4B ) . This indicates that Fkh1 does not bind non-functional replication origins and suggests that recruitment of Fkh1 to early origins may be coupled with origin licensing . Licensing of replication origins begins with the recruitment of ORC to the ACS motif and is completed during G1 phase by Cdc6- and Cdt1-mediated loading of MCM double hexamer complex onto the origins [9 , 10] . In addition , several early firing origins are also pre-loaded with the Cdc45 protein that is required during the subsequent S phase for activation of MCM helicase [11 , 12] . To determine which step of origin licensing is necessary for efficient binding of Fkh1 to origins , we monitored its recruitment to the GAL-VPS13-ARS607 locus in a re-licensing assay in a set of strains expressing temperature sensitive mutants of Cdc6 , Mcm2 or Cdc45 proteins . The re-licensing assay using GAL-VPS13-ARS607 is based on the fact that all pre-RC components can be removed by active transcription over the replication origin and , if the cell remains continuously arrested in G1 , these origins become re-licensed rapidly upon shut-down of transcription [13] . The general outline of the assay is shown in Fig 4C . First , cells were arrested in G1 and kept arrested throughout the rest of the experiment . Next , transcription of GAL-VPS13-ARS607 was activated , leading to displacement of all pre-RC components and Fkh1 proteins from the locus . At the same time , the cells were shifted to a non-permissive temperature to inactivate Cdc6 , Mcm2 , or Cdc45 proteins . After two hours of incubation , transcription of GAL-VPS13-ARS607 was repressed , which in turn enabled re-licensing of the locus up to the step where the temperature-sensitive component of the pathway was required . As expected , MCM was reloaded to the locus in wt and cdc45-ts strains , but not in cdc6-ts and mcm2-ts strains ( Fig 4D ) . Similarly , Fkh1 failed to rebind the origin in cdc6-ts and mcm2-ts strains , while it was efficiently re-recruited in wt and cdc45-ts strains ( Fig 4E ) , indicating that Fkh1 was recruited to origins during replication licensing together with or shortly after the loading of MCM complex . These results also suggest that Fkh1 binds replication origins in cell cycle dependent manner , i . e . in G1 when pre-RCs are present , but not in G2/M phase when DNA replication is finished and new pre-RCs are not formed yet . To confirm this , we compared the recruitment of Fkh1 to origins in G1 and M phases of the cell cycle . As expected from earlier results , Fkh1 was bound to ARS305 , ARS607 , ARS737 and VPS13-ARS305 in G1 , but not in M phase , while Fkh1 binding to its recognition sequence in the CLB2 promoter was not affected by the cell cycle ( Fig 4F ) .
In eukaryotic cells , DNA replication is initiated from numerous origins throughout the S phase . Temporal activation of origins is regulated by a variety of factors including the origin’s location and local chromatin context . However , some origins appear to be immune to effects of local chromatin structures and fire early even when transposed to new naturally late-replicating genomic loci . Recent studies have shown that Forkhead family transcription factors are responsible for early activation of many origins . Forkhead-regulated origins typically contain multiple Fkh1/2 consensus binding sites near the ACS , and at least two of these are required for their early activation [3 , 4] . Interestingly , at origins where Fkh1 binding was studied in greater detail ( ARS305 , ARS607 and ARS737 ) , the consensus sequences are arranged in an identical pattern: the two sites are oriented divergently and separated by 72 base pairs . Additionally , one of the sites is located very close to the ACS , overlapping it at ARS607 and ARS737 ( Fig 4A and S3 Fig ) . In order to determine whether the exact spacing and orientation of Fkh1/2 sites are critical for Fkh1 recruitment and early activation of these origins , we constructed a panel of yeast strains with altered Fkh1/2 consensus site configuration . At ARS607 , we altered the spacing between the two sites , while at ARS305 we modified their relative orientation . Consensus site reversal could not be done in ARS607 as the proximal site overlaps the ACS and its rotation would have inactivated the origin . On the other hand , multiple Fkh1/2 consensus sites present near the ARS305 locus left very limited possibilities to change the distance between the two key sites at this origin . By contrast , ARS607 has no other Fkh1/2 sites within 400 bp of the ACS in 3’ direction . To avoid the possible influence of other DNA replication origins and Fkh1/2 consensus sites near their native loci , the modified ARS305 and ARS607 were inserted into the ectopic naturally late-replicating GAL-VPS13 locus . We have shown previously that both origins are fully functional and fire early in GAL-VPS13 if their two Fkh1/2 consensus sites remain intact [3 , 13] . Our results show that Fkh1/2 sites are very precisely arranged near the early replication origins and that no alterations in their configuration are tolerated . Changing the spacing between the sites or reversing their orientation leads to significant decreases in Fkh1 binding to the origins ( Fig 1C and 2B ) . Accordingly , such modified origins also fail to fire early in the S phase ( Fig 1D and 2C ) . The only exception to this general pattern was observed when the distance between the two sites was reduced to 10 bp . However , while this change had only a modest effect on Fkh1 binding , the altered origin failed to fire early in S phase . This observation underscores the delicate nature of the mechanism behind Fkh1’s role in replication regulation , indicating that mere binding of this protein is insufficient for proper regulation of an origin’s firing time . Our findings are also supported by genome-wide replication initiation data . We observed that the 72 base-pair gap and divergent orientation of Fkh1/2 sites was present at 48 locations throughout the budding yeast genome , with 7 of those overlapping early-firing origins ( S1 Table ) . Relaxing the search criteria by allowing a 71–73 bp gap between the sites increased the number of total hits but did not change the fact that divergent Fkh1/2 consensus sequences preferentially co-localized with early replicating origins ( Fig 3A ) . Combined data from our genome-wide analysis of replication origins revealed that divergently oriented Fkh1/2 consensus sites separated by 71–79 bp overlapped with early origins more frequently than was expected from random distribution of such sites . However , this result may overestimate the tolerance in the gap size between Fkh1/2 sites , given our finding that increasing the gap from 72 bp to 77 bp at ARS607 results in loss of its ability to bind Fkh1 and to fire early in S phase ( Fig 1C and 1D ) . Overall , our results with modified ARS607 and ARS305 indicate that even minor rearrangements of the Fkh1/2 sites are not tolerated and suggest that binding of Fkh1 to replication origins requires precise arrangement of Fkh1/2 binding sites in the locus . Previous studies have revealed that several sequence elements within ARS305 and ARS607 are essential for full activity of these origins . Deletion analysis of ARS305 demonstrated that the region including the ACS-proximal Fkh1/2 site was required for the origin’s function [14] . Moreover , systematic mutational analysis of ARS305 identified three short regions , in addition to the ACS , that influenced the stability of plasmids carrying this ARS as a sole replication origin . Mutation of nucleotides immediately adjacent to the 11-bp ACS lead to a complete loss of the origin’s activity , while disruption of either Fkh1/2 site lead to a significant decrease in plasmid retention during exponential cell growth [15] . In addition , yeast DNA replication origins contain nuclease hypersensitive A/T-rich sequences near the ACS , termed DNA unwinding elements [16] . In early origins , these sequences contain one or several continuous stretches of adenine nucleotides ( S3 Fig ) that may be required for efficient opening of DNA strands by the MCM helicase . Disruption of Fkh1/2 sites or A-tracks within ARS607 leads to decreased mitotic stability of plasmids , demonstrating the contribution made by these elements to the full activity of the origin [17 , 18] . These results support the view that Forkhead proteins enhance replication origins’ efficiency by marking them as ‘first to fire’ when DNA synthesis begins . If Forkhead binding is disturbed , the origin remains functional but loses its early-firing properties and concordantly , becomes less efficient . Complexity of Fkh1 binding to DNA was further supported by the discovery that formation of the pre-RC at origins was necessary for efficient recruitment of Fkh1 . Fkh1 binding to ACS-mutated origins was severely reduced despite the fact that all such loci contained correctly oriented and spaced Fkh1/2 sites ( Fig 4B ) . This suggests either that Fkh1 binding is enhanced by interactions with pre-RC components , or that pre-RC-directed chromatin reorganisation is required for Fkh1 to gain access to the Fkh1/2 consensus sites . Previous studies have shown that replication origins are flanked by strongly positioned nucleosomes and that the ORC complex is needed for this arrangement in vivo and in vitro [19 , 20] . Therefore , accessibility of Fkh1/2 sites may be compromised without ORC-dependent nucleosome positioning . On the other hand , one Fkh1/2 site often overlaps the ACS in early origins , suggesting that binding of the ORC and Forkhead are mutually exclusive . However , recent in vitro studies demonstrate that in addition to the ACS , ORC also binds other ACS-like sequences within origins , and that its selectivity towards different binding sites is partially regulated by its interaction with Cdc6 during origin licensing . Moreover , after MCM loading is completed , ORC is removed from the ACS [21 , 22] . These results suggest that the ACS is an essential entry point for sequential loading of different pre-RC components onto origins . However , it is not necessarily the final binding site for all recruited proteins complexes . Therefore , redistribution of pre-RC components around the ACS may provide an opportunity for the Forkhead proteins to bind their recognition site within the ACS . To determine which steps in pre-RC formation are critical for Fkh1 recruitment , we used the origin re-licensing assay [13] that allowed us to monitor re-binding of Fkh1 to the origin in conditions where different pre-RC components were inactivated by temperature-sensitive mutations ( Fig 4C ) . We observed that Fkh1 was not recruited to the origin in cdc6-ts or mcm2-ts strains , while it was successfully reloaded in a cdc45-ts strain ( Fig 4E ) . This indicates that Fkh1 is recruited to origins at the same time or shortly after the pre-RC is fully formed and the Mcm2-7 complex is loaded . This model was further supported by the observation that Fkh1 occupancy at origins decreases significantly in M phase , where origins are not licensed ( Fig 4F ) . By contrast , the next step–recruitment of Cdc45 –is not required for Fkh1 binding . Once recruited to origins , the binding of Fkh1 is presumably stabilised by the ORC complex , as Forkheads interact directly with ORC proteins [4] . However , apparently the ORC alone is not sufficient for efficient recruitment of Forkheads , as the formation of entire pre-RC is required for successful binding of Fkh1 to the origin ( Fig 4E ) . Interestingly , the 71–79 bp gap between Fkh1/2 sites in early origins corresponds very closely to the footprint of the Mcm2-7 double hexamer , which covers about 70–80 bp DNA when loaded onto origins [23 , 24] . Therefore , it is possible that on licensed origins the Mcm2-7 complex is stabilised by Forkhead proteins that flank the helicase on both sides . Reciprocally , loading of the Mcm2-7 complex may be necessary to fully expose the Fkh1/2 binding sites . We also noticed that in several early origins one of the Fkh1/2 sites is ‘doubled’–it contains two consensus sequences that overlap partially ( S3 Fig ) . This provides some flexibility of the gap size between Fkh1/2 sites , which might help fine-tune the loading of the pre-RC and Forkhead proteins . Overall , these results suggest that binding of Fkh1 to replication origins , and possibly to other genomic locations , is a finely regulated process that requires precise arrangement of Fkh1/2 binding sites and the presence of supporting protein complexes in the locus .
All Saccharomyces cerevisiae strains were congenic with strain W303 and are listed in S2 Table . The GAL-VPS13-ARS strains contain different versions of ARS305 and ARS607 in the GAL-VPS13 locus , at 3220 bp downstream from the VPS13 start codon . The following ARS sequences were used for construction of GAL-VPS13-ARS strains ( sequence coordinates from the Saccharomyces Genome Database , http://www . yeastgenome . org ) : ARS305 ( Chr3 , nucleotides 39529–39800 ) ; ARS607 ( Chr6 , nucleotides 199392–199779 ) . To change the distance between Fkh1/2 binding sites in GAL-VPS13-ARS607 , the distal Fkh1/2 site was mutated ( GTAAATA to GATCCTA ) and then a new Fkh1/2 site ( GTAAATA ) was inserted at various distances ( 10 , 30 , 60 , 90 , 120 , 150 , 180 , 240 , or 300 bp ) away from the ACS-proximal Fkh1/2 site . For finely mapping the tolerance of Fkh binding for altered gap size between Fkh1/2 binding sites within ARS607 , insertions of 5 , 10 , or 15 bp were introduced between the two sites in GAL-VPS13-ARS607 locus . All insertions were located at a distance of 7 bp from the distal Fkh1/2 binding site . 10 bp deletions between the Fkh1/2 sites were made in two different positions in ARS607 , one located 24 bp and the other 2 bp away from the distal Fkh1/2 site . In GAL-VPS13-ARS305 , one or both Fkh1/2 consensus binding sequences were reversed ( 5’ site: TGTTTAT to ATAAACA; 3’ site: GTAAATA to TATTTAC ) . In strains AKY956 and AKY952 , ACS of GAL-VPS13-ARS305 , or GAL-VPS13-ARS607 was mutated ( in ARS305: TTTATATGTTTT to TTTATATGggTT; in ARS607: GTTTATATTTAG to GTTTATATccAG ) . ACS of ARS305 and ARS737 ( TTTTAATATTT to TTTTAATAccc ) were mutated in their native loci in strains AKY1121 and AKY1122 , respectively . Sequences of all modified origins are shown in S4 Fig . All modified origins were inserted into genomic loci by two step gene replacement protocol . First , URA3 gene was inserted into the desired locus and then replaced with ARS sequence by homologous recombination and counter-selection on 5-FOA plates . Strains carrying temperature sensitive alleles cdc6-1 , mcm2-td , or cdc45-td [25–27] were used to construct strains AKY1061 , AKY1143 and AKY1144 for the origin re-licensing assay . For efficient α-factor arrest , the BAR1 gene was deleted in all strains . For ChIP assays , the Fkh1 protein was tagged with C-terminal triple E2-tag recognized by a 5E11 antibody ( Icosagen ) , while Mcm4 was tagged with C-terminal triple myc-tag recognized by a 9E10 antibody . Cells were grown in yeast extract-peptone-dextrose ( YPD ) medium containing 2% glucose as a carbon source before fixation with 1% formaldehyde for the ChIP assay . Cell cycle arrest in G1 was achieved by addition of α-factor-mating pheromone ( Zymo Research ) to the growth media to a final concentration of 100 nM and by further incubation for 3 hours . Cell cycle arrest in M phase was achieved by incubating the cells with nocadazole ( Sigma ) with final concentration 20 μg/ml for 60 minutes . ChIP assays were performed as described previously [28] . Shortly , whole-cell extract from 107 cells was used for ChIP assays with 0 . 5 μg of anti-E2 tag antibody ( 5E11 ) or 1 μg of anti myc-tag antibody ( 9E10 ) . Co-precipitated DNA was analysed by quantitative PCR ( qPCR ) using Roche Lightcycler 480 real-time PCR system under standard conditions ( 40 cycles; 95°C for 15 s , 58°C for 20 s , 72°C for 20 s ) . Maxima SYBR Green/ROX qPCR Master Mix ( Thermo Scientific ) was used . qPCR was done with primer pairs covering the relevant regions of VPS13 as well as native origins ARS607 , ARS305 , ARS737 and ARS522 . Signals were normalized to the high copy-number telomeric PAU1 gene ( in GAL-VPS13-ARS607 strains to the native ARS607 origin ) . Presented results show the average of three independent experiments , error bars indicate standard deviations . Sequences of qPCR primers are shown in S3 Table . Yeast strains were arrested in G1 for 3 hours with α-factor and then released into YPD media containing 200mM hydroxyurea ( HU ) at 24°C . Samples were collected 45 and 75 minutes later by fixing approximately 2x107 cells in 80% ethanol on ice . Subsequently , cells were washed twice with water and disrupted with 0 . 5mm glass beads in lysis buffer ( 2% Triton X-100 , 1% SDS , 100 mM NaCl , 10 mM Tris-HCl pH 8 . 0 , 1 mM EDTA pH 8 . 0 ) . Cell lysate was incubated at 40°C for 15 min , genomic DNA was extracted with phenol-chloroform , precipitated with ethanol and dissolved in water . The relative amount of DNA was determined by qPCR with primers specific for ARS305 , ARS607 , ARS522 and VPS13 loci . The late-replicating locus ARS522 was used to normalize the data and to calculate the relative increase of the DNA amount at other loci during the experiment . Presented results show the average of three independent experiments , error bars indicate standard deviations . Yeast strains were grown at 24°C in YP-raffinose media for up to 48 hours to obtain culture densities of approximately 1×107 cells/ml . Cells were arrested in G1 for 3 hours with α-factor , then washed once with water and transferred to YP-galactose ( containing α-factor ) to induce transcription of the GAL-VPS13-ARS607 cassette . Additionally , during the galactose treatment , temperature was shifted to 37°C to activate the degron system utilized to inactivate Mcm2 or Cdc45 proteins . In order to ensure destruction of Mcm2 and Cdc45 , as well as of the temperature-sensitive version of Cdc6 , all subsequent steps were carried out at 37°C . Cultures were grown in YP-galactose for 2 hours , following which cells were washed with water and transferred to YPD ( pre-warmed to 37°C ) containing α-factor . Cultures were incubated at 37°C for 40 minutes , samples were cross-linked with formaldehyde and processed for ChIP analysis . ChIP data were normalized on the native GAL10 gene , which was regulated by carbon source changes in parallel with the GAL-VPS13-ARS607 cassette . Saccharomyces cerevisiae genome was scanned for single and double Fkh1/2 consensus binding sites ( RYMAAYA ) . For tandem sequences , all possible orientations of the sites: divergent ( ‘head-to-head’ ) , convergent ( ‘tail-to-tail’ ) , unidirectional ( ‘head-to-tail’ ) were included and gaps of 62 to 88 bp between the sites were allowed . Midpoint coordinates of discovered double Forkhead binding motifs were plotted against the genome-wide dataset of early DNA replication initiation profile ( as determined by BrdU incorporation in the presence of HU ) [8] . On each chromosome , the highest BrdU signal that did not overlap with any of the confirmed replication origins was set as threshold value . All BrdU peaks over the threshold value were considered to represent genuine early replication origins . Midpoint coordinates of double Forkhead binding sites that were found within 200 bp from BrdU peaks maximum signal were considered as overlapping with the origin . Random overlap was calculated as average overlap between the peaks of early origins and scrambled Fkh1/2 consensus sequences ( YAAYMAR , MAARYAY , AAYMYAR ) in all orientations , separated by 50–100 bp . All combinations of scrambled sequences were found uniformly over the replication origins with no preference for any particular sequence or orientation ( S5 Fig ) . Late origins were defined as confirmed DNA replication origins in the S . cerevisiae OriDB ( http://cerevisiae . oridb . org ) [29] with defined ARS sequence no longer than 600 bp that did not initiate replication in the presence of HU [8] . The analysis of Fkh1/2 consensus site distribution in late origins was performed using the same methodology as for early origins .
|
In this study , we explore the mechanisms that determine activation of DNA replication origins in early S phase . It has been shown that a subset of replication origins is regulated by Forkhead family transcription factors that ensure their firing at the beginning of S phase . However , the recruitment of Forkhead factors to replication origins is not a straightforward process–there are thousands of Forkhead binding sites in the genome and their presence does not guarantee that Forkheads actually bind these sites . We show that recruitment of Fkh1 protein to DNA replication origins requires precise arrangement of Forkhead binding sites and depends on formation of pre-replicative complexes at the origins . These results clarify the mechanisms of Forkhead-dependent regulation of early DNA replication origins and also reveal that mere presence of consensus binding sites is not sufficient for recruitment of Forkhead proteins to their target loci .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"&",
"methods"
] |
[
"insertion",
"mutation",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"dna-binding",
"proteins",
"mutation",
"fungi",
"model",
"organisms",
"dna",
"replication",
"experimental",
"organism",
"systems",
"sequence",
"motif",
"analysis",
"dna",
"synthesis",
"phase",
"research",
"and",
"analysis",
"methods",
"saccharomyces",
"sequence",
"analysis",
"bioinformatics",
"proteins",
"genetic",
"loci",
"yeast",
"biochemistry",
"cell",
"biology",
"nucleic",
"acids",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"organisms"
] |
2017
|
Recruitment of Fkh1 to replication origins requires precisely positioned Fkh1/2 binding sites and concurrent assembly of the pre-replicative complex
|
Leptospirosis is an important cause of seasonal outbreaks in New Caledonia and the tropics . Using time series derived from high-quality laboratory-based surveillance from 2000–2012 , we evaluated whether climatic factors , including El Niño Southern Oscillation ( ENSO ) and meteorological conditions allow for the prediction of leptospirosis outbreaks in New Caledonia . We found that La Niña periods are associated with high rainfall , and both of these factors were in turn , temporally associated with outbreaks of leptospirosis . The sea surface temperature in El Niño Box 4 allowed forecasting of leptospirosis outbreaks four months into the future , a time lag allowing public health authorities to increase preparedness . To our knowledge , our observations in New Caledonia are the first demonstration that ENSO has a strong association with leptospirosis . This association should be tested in other regions in the South Pacific , Asia or Latin America where ENSO may drive climate variability and the risk for leptospirosis outbreaks .
Leptospirosis is an important zoonotic disease with high incidence in tropical and subtropical regions worldwide , with more than 850 , 000 cases a year , a high incidence in tropical regions worldwide [1] and a case fatality ratio frequently exceeding 10% [2] . It has also been reported as an emerging or re-emerging disease , including in temperate countries [3] , [4] . More than 230 pathogenic serovars have been described , belonging to 9 pathogenic and 5 intermediate Leptospira species [5] . Humans acquire infection through direct contact with the urine or kidney tissues of infected reservoir mammals or , most frequently , through contaminated water [6] , [7] . Because of this transmission routes , climate is a major driver of leptospirosis . Increased rainfall leads to increased human exposure [8] through both increased survival of the bacteria in the environment and increased exposure of humans to surface water [9] . Extreme climatic events and floods have frequently been associated with leptospirosis outbreaks [10]–[12] . Rainfall also leads to larger rodent populations , further contributing to increased environmental contamination [13] . El Niño events occur when a warm ocean water pool occasionally reaches the Pacific coast of Latin America . The opposite phase is called “La Niña” and refers to the movement of this warm ocean water pool westwards . This oceanic phenomenon is strongly linked to atmospheric pressure changes known as the “Southern Oscillation , ” and their interplay is usually called El Niño Southern Oscillation or ENSO . This interaction between ocean temperatures and pressure is associated with increased rainfall , including heavy rain episodes and floods . Conversely , droughts can affect the opposite side of the South Pacific ( e . g . Australia ) during El Niño episodes . This phenomenon is quasi-periodic and occurs every two to seven years . Because ENSO is an important determinant of year-to-year variability in weather , including heavy rainfall and drought , a number of studies have focused on its impact on transmission and disease patterns . ENSO has been associated with outbreaks of cholera and a range of vector-borne diseases including leishmaniasis , malaria , and arboviral diseases in various countries of Latin America and Asia [14] , [15] . More recently , together with rodent density , ENSO has also been associated with hantaviruses incidence in China [16] . Surprisingly , despite its major impact on rainfall , investigations have not clearly demonstrated a link between ENSO and leptospirosis , though it has been suspected to be related to a four-fold increase in the incidence of leptospirosis in Guadeloupe from 2002 to 2004 [17] . New Caledonia , a French territory in the Southwest Pacific ( 20–22°S , 164–167°E ) , provides a unique setting to understand the epidemiology of leptospirosis , a major public health concern for the island [18]–[20]and its link to climatic variability . Notification of leptospirosis to the health authority has been mandatory since 1991 . Furthermore surveillance is laboratory-based , where a single reference laboratory has been performing laboratory confirmation of the diagnosis since 1989 [19] , [21] , [22] . This system has yielded accurate data on human leptospirosis incidence , which occasionally exceeds 100 cases per 100 000 inhabitants a year . Leptospirosis cases in New Caledonia are most common among people living in tribes or villages in the rural backcountry , though it also affects a few urban citizens . The disease is strongly seasonal , with epidemics occurring during the hot and rainy season from January through June each year . However , strong variations in incidence occur between years , epidemics occurring during periods of heavy rains [23] . Although suspected [18] , the link between ENSO and year-to-year variation in the size of seasonal outbreaks has not been investigated with extended time series data from New Caledonia . Furthermore an improved understanding of how climatic and meteorological factors contribute to epidemics may allow for early-warning outbreak predictions and implementation of more effective public health preparedness and response . In this study , we aimed to determine whether ENSO-related variations in climate and meteorological conditions are associated with leptospirosis epidemics in New Caledonia and to determine whether such data could be used to build an early-warning system .
For leptospirosis diagnosis , serum and/or urine ( and occasionally other sources , like cerebrospinal fluid ) specimens are routinely received at Institut Pasteur laboratory from all hospitals , health centers and private laboratories throughout New Caledonia . As recommended by the World Health Organization [24] , serology is carried out using the reference Microscopic Agglutination Test ( MAT ) with a 11-strain panel suited to the local epidemiology [18] . The microbiological diagnosis relied on culture and a nested PCR targeting the 16SrRNA gene [25] prior to 2006 and then shifted to a real time PCR specific for pathogenic leptospires . From 2006 to 2011 , the lfb1 gene was targeted using SYBR Green I technology [26]; the technique was then changed in 2012 for a TaqMan probe-based technique targeting the pathogenic Leptospira-specific lipL32 gene [27] . Both techniques have a similar lower limit of detection [28] , a finding confirmed during our technique transition period ( data not shown ) . The case definitions used for leptospirosis surveillance in New Caledonia have been the same since 1995 and were described previously [18] . Briefly , based on a clinical suspicion , cases were considered as confirmed if Leptospira is cultured , or if its genome is evidenced by PCR or real time PCR from blood , urine or cerebrospinal fluid , or if a seroconversion ( from nil to at least a 400 titer ) or a significant sero-ascension ( at least a fourfold raise in titers ) is observed in paired sera using the MAT . The case was considered as probable leptospirosis if a single MAT titer ≥800 is observed for at least one pathogenic serogroup in the serum of a patient with a clinical suspicion . Anonymous data used in this study were extracted without patient identification from the Institut Pasteur laboratory database and originated from routine diagnostic activities as part of public health surveillance . The corresponding biobank was declared to the French Ministry of Research ( DC-2010-1222 , Collections number 1 and 2 ) . We excluded patients from other Pacific Islands Countries or Territories whose specimens are sometimes submitted to Institut Pasteur laboratory . The date assigned to each case was the month when the first biological specimen was collected for submission to the laboratory ( e . g . date of the first serum in the case of paired sera , even if MAT-negative ) . The ENSO is studied using both atmospheric and oceanographic parameters . The Southern Oscillation Index ( SOI ) reflects the atmospheric pressure difference between Darwin in Australia and Tahiti in French Polynesia [29] . A sustained value of the SOI under −8 ( or above +8 respectively ) is considered as reflecting an El Niño ( or a La Niña respectively ) status of the oscillation . Oceanographic parameters include Sea Surface Temperatures ( SST ) and their anomalies [30] in four equatorial oceanic “Boxes” , Boxes 1 and 2 being adjacent to South American Pacific coasts and Box 4 being the most westward one . An El Niño phase of the ENSO is related to positive SST anomalies in the Niño Box 4 . We studied SST in the most commonly used region 3 . 4 and in the Box 4 , the most relevant area from a geographical point of view . Other indices have been proposed and used including the Oceanic Niño Index ( ONI ) [31] and the Multivariate El Niño Index ( MEI ) that combines sea-level pressure , components of the surface wind , SST , surface air temperature and cloudiness indicators [32] , [33] . The study period was January 2000–December 2012 for leptospirosis cases and July 1999–December 2012 for oceanic and atmospheric ENSO parameters . Climatic and oceanographic data were downloaded from the Australian Bureau of Meteorology ( Southern Oscillation Index , SOI ) or the National Oceanic and Atmospheric Administration ( NOAA ) websites ( Sea Surface Temperatures and anomalies , ONI and MEI ) . Meteorological data for New Caledonia were chosen from 3 leptospirosis hot spots located in the middle of the country [23] , all kindly provided by Meteo France . A map displaying the location of the meteorological stations is provided as Figure S1 as Supporting Information . We first tested for an association between monthly cases of leptospirosis and each of the El Niño variables ( surface temperature , El Niño indices ) and climate variables ( average monthly minimum/maximum temperature , cumulative rainfall over the previous 1–8 months . These associations were evaluated using negative binomial regression because the data exhibited evidence of overdispersion . All of the models included harmonic terms ( sine and cosine terms that repeat every 12 months ) to control for consistent , shared seasonal variations that were unrelated to variations in the climate variable . The baseline model included the harmonic terms but did not include any climate or meteorological variables . In addition to the climate and meteorological variables , we tested whether the number of cases in month t-1 or t-12 predicted the number of cases in the current month ( t ) ( log-transformed variable , adding 0 . 5 to each observation ) . Since the effect of climate on disease might not be immediate , we evaluated associations between leptospirosis cases and each of the El Niño and climate variables with lags of 0 to 6 months . In total , 46 different variables were tested with 7 lags each ( 0 to 6 months ) for a total of 322 single variable models . Bayesian information criteria ( BIC ) were used to identify the variables that best explained the disease data . We also evaluated the correlations among the climate and meteorological variables using partial correlations ( PROC CORR in SAS v9 . 3 ) , controlling for harmonic variation with a sine and cosine variable with a 12 month harmonic . To estimate the total contribution of climate and meteorological variation to the burden of leptospirosis , we next performed multivariate analysis . Because there was a high degree of collinearity among the potential climate and meteorological variables [34] , we reduced the number of variables in the model by first performing principal components analysis ( PROC PRINCOMP in SAS v9 . 3 ) [35] . The first 5 principal components accounted for 81 . 3% of the variation in the climate and meteorological variables ( Supporting Table S1 ) . We then used the BMA package in R [36] , [37] to perform BIC-based model averaging [38] . The variables for model averaging included the first five principal components [35] , sine and cosine terms with a 12 month period , and 1-month , and 12-month autoregressive terms ( log-transformed ) . The outcome variable was monthly leptospirosis cases . The number of leptospirosis cases attributable to each of the principal components was determined by: predicted- ( predicted/exp ( βxi*xi ) ) , which compares the total number of predicted cases in each month with the number of predicted cases when the component ( βxi ) is held to 0 . The attributable percent is calculated as the number of cases attributable to each component divided by the total number of predicted cases . This value can be interpreted as the percent of cases that would not have occurred if the specific factor was held at its mean ( no variation ) . We considered whether outbreaks of leptospirosis could be forecast several months in advance . The goal was to develop a simple model based on a small number of components that are available in a timely manner . Based on the univariate analyses , we decided to use sea surface temperature anomalies ( box 4 , lagged by 4 months ) as the primary prediction variable . SST anomaly ( box 4 ) was among the best predictors of leptospirosis cases based on BIC score and is available more rapidly than the ONI index . We also included sine and cosine terms ( 12 month period ) , and the number of cases 12-months prior to the month being forecast ( log-transform ) . The initial training period for fitting the model was from 2000–2006 , and we iteratively added an additional month onto the training period and refit the model . For each of these training periods , we fit the model and then extrapolated the number of cases 4 months after the end of the training period . The observed number of cases 4 months past the training period was compared to the predicted number of cases using correlations . We also considered whether these models could accurately detect whether an epidemic would occur ( binary for the number of cases being above or below an epidemic threshold ) . An epidemic threshold was estimated using the Serfling approach [39] , where a harmonic baseline was fit to the square root-transformed leptospirosis data from June–September from 2000–2012 . An outbreak was defined according to the forecasting model if the observed or expected number of cases exceeded the upper 95% prediction interval .
A total of 1163 ( 731 confirmed and 432 probable ) human leptospirosis cases were diagnosed over the period 2000–2012 . The age range was 1–84 . 8 year old ( out of 729 archived data ) , the median age was 34 . 0 . Most cases were males ( Sex-ratio Male/Female = 1 . 90 out of 1150 data available ) . During this same period , leptospirosis was considered by New Caledonian Health Authority as the cause for 40 deaths , varying between one ( in 2007 and 2010 ) and 7 deaths per year ( in 2001 ) , a mean 3 . 4% fatality rate over the study period [40] . During the surveillance period , the mean annual incidence was 37 . 4 cases per 100 , 000 population . The number of cases however was highly variable ( Figure 1A ) and as a consequence , the incidence fluctuated between 5 . 6 ( in 2004 ) and 66 . 4 ( in 2009 ) cases per 100 , 000 population . During our study period significant variations occurred in the SOI and SST anomalies in the oceanographic Niño Box 4 ( Figure 1C ) . There were 4 El Niño and 5 La Niña periods from 2000–2012 . There were moderate but significant negative associations between the El Niño indices and rainfall and temperature measurements ( Supporting Table S2 ) . There was a significant association between leptospirosis cases and each of the El Niño indices , sea surface temperature anomalies , and rainfall ( Figure 1B ) , even when controlling for consistent shared seasonal variations with harmonic variables ( Table 1 , Supporting Table S2 ) . In particular , the Oceanic Niño Index and SST anomaly ( box 4 ) best fit the leptospirosis data ( Table 1 , Supporting Table S3 ) . The association between leptospirosis and SST anomaly was greatest with a 4 month lag ( cases increased 4 months after the change in SST ) ( Table S2 ) . The meteorological variable with the best fit to the leptospirosis data was the cumulative rainfall in Poindimie over the previous 8 months ( Supporting Table S3 ) . We next used model averaging to fit a multivariate model that included the climate and meteorological variables ( represented by 5 principal components ) , harmonic terms , and autoregressive terms . While the associations between leptospirosis and individual principal components are not meaningful in this context , the full averaged model accurately captured the dynamics of leptospirosis cases , including the large epidemics and the series of weaker years in the mid-2000s ( Figure 2 ) . Based on the model fit , we estimated that 35 . 8% of leptospirosis cases were attributable to variability in the climate and meteorological variables . Finally , we considered whether we could forecast the number of leptospirosis cases four months into the future using a simple model that included SST anomaly in Niño Box 4 , which was among the best predictors of leptospirosis in the single variable analysis and is available nearly in real-time . The model included the SST anomaly variable ( Niño Box 4 , lagged by 4 months ) , the number of leptospirosis lagged by 12-months and the harmonic variables to forecast the number of cases four months ahead . There was a moderate-to-strong correlation between the observed and forecasted values ( Spearman's r = 0 . 74 ) . The observed number of leptospirosis cases exceeded an epidemic threshold in 2008 , 2009 and 2011 . The model correctly predicted that the number of cases would exceed the threshold in each of these years ( Figure 3 ) . Likewise , the model correctly predicted the 2007 and 2010 would be non-epidemic years . The model incorrectly predicted that the observed number of cases would exceed the epidemic threshold in 2012 ( Figure 3 ) . For the 2008 and 2011 epidemics , the model correctly predicted the first month that would be above the threshold ( Figure 3 ) . In contrast , in 2009 , the first month that was predicted to be above the threshold was two months after the actual start of the epidemic . Of note , the 4-month-lagged climate variable still would have predicted an epidemic two months in advance of the actual start of the 2009 epidemic .
Leptospirosis is a major infectious disease in Pacific Island Countries and Territories ( PICTs ) [41] . It usually has a rural endemic pattern with seasonal oscillations in incidence and strong inter-annual variability in incidence . Here we demonstrate that ENSO-associated variations in SST , and related variations in rainfall and temperature , are associated with large variations in leptospirosis incidence in New Caledonia . La Niña periods were previously shown to be responsible for heavy rainfall in New Caledonia [42] , [43] . Mechanistically , there is strong support for the notion that climate can influence leptospirosis incidence . Apart from isolated extreme rainfall events , known to be possible triggers of leptospirosis outbreaks [12] , [44] , [45] , our results suggest a more complex and longer lasting mechanism . The single variable analyses show that the accumulated rainfall during the previous 8 months is linked to leptospirosis incidence . One possible explanation is that periods of intense rainfall build-up the rodent reservoir , both by increasing the density of rodent populations and the prevalence of Leptospira within these populations [13] , which would in turn increase environmental contamination . In the environment itself , wet and hot conditions will also favor Leptospira survival , also contributing to increased exposure in humans [13] . During these periods , humans are also more likely to be exposed to floods or flooded grounds , again increasing exposure risk . Similarly , by facilitating transmission between animals , large domestic or feral mammals like cattle , deer and pigs , could have an increased carrier rate after months of humid conditions . Therefore , large mammals probably also contribute to the transmission over consecutive years . The hypothesis of an increased capacity of the animal reservoir supported by both rodents and large Mammals is notably supported by a significant one-year lag impact of the incidence , a duration not explained by the abundance of short-lived rodents . We found the strongest link between SST anomalies and leptospirosis cases , but we also found that variations in SST influenced rainfall and temperature , which in turn were directly associated with the risk for leptospirosis . Pathogenic Leptospira typically survive in wet soils , and heavy rain storms can flush the bacteria out into the water bodies where they can more readily infect humans . Therefore , there is a plausible causal link between that the changes in SST influence rainfall and leptospirosis cases . More generally , ENSO has a major impact on rainfall inter-annual variability in many regions in the world . The results of this study should be evaluated in other countries in the Pacific , South Asia , Latin America , or Eastern Australia . New Caledonia is unique in its island geography , and New Caledonia's weather variability is highly impacted by ENSO . Therefore , other studies will need to determine whether SST anomalies or other ENSO-related parameters are good predictors of rainfall and leptospirosis cases in other , inland regions or island territories . In large countries like Thailand , there is evidence to suggest that different models should be used for different regions [46] , making country-wide preparedness difficult . In the smaller Reunion Island [47] , a season and meteorological-based model provides a good description of the incidence of the disease and might be used for prediction with a satisfactory accuracy . Our study has limitations . Meteorological data were available from several weather stations around New Caledonia , but it was uncertain which station would best reflect the exposures of the population . We therefore evaluated several stations chosen in areas of highest leptospirosis incidence [23] . This is an ecological study , and we did not directly measure the impact of SST or rainfall on rodent density or Leptospira exposure in the environment . However , this link is well-established , and there is strong plausibility for a meteorological link for leptospirosis . The choice of an epidemic threshold is an arbitrary choice that should be driven by public health necessity . We used a seasonal “Serfling” baseline with a 95% confidence interval , which is commonly used for detecting influenza outbreaks . More or less conservative baselines could also be appropriate depending on the objective . In these exploratory analyses , we considered many different El Niño and meteorological variables , with various lags . To reduce the number of predictors and the impacts of collinearity , we performed principal components analysis priori to multivariate model averaging [35] . While this analysis clearly established a strong correlation between the climate/meteorological variables and leptospirosis cases , the results should be interpreted cautiously . Since rainfall and temperature are likely on the causal pathway between El Niño and leptospirosis , including the meteorological variables in the multivariate model dampens the association between El Niño and leptospirosis . The “forecasting” analysis was performed after SST anomaly was identified as being important in the full dataset . Additional years of data will be required to confirm the usefulness of this model in forecasting epidemics in New Caledonia . Predicting outbreaks of leptospirosis is of prime importance for public health decision makers . Predictive models for leptospirosis are scarce . In our study , we demonstrate that the SST anomaly in the Niño Box 4 would allow a good prediction of leptospirosis incidence in New Caledonia 4 months in advance . This lag could allow increasing awareness and preparedness at all levels , from the general population to health practitioners . The use of mass media ( radio or TV spots , advertising posters ) for general population could allow reminding of risk and exposition factors as well as symptoms and encourage to promptly seeking medical advice . Likewise , medical professional forums and meetings could be used to increase awareness in the medical community . Because some at-risk occupations are vaccinated in New Caledonia , booster injections could also be planned and administered as part of this increased preparedness . Because particularly timely , it would probably additionally allow implementing practical field prevention measures , like rodent control operations or the cleaning out of river banks and sewage systems to minimize flooding risks , targeting areas of highest incidence .
|
The El Niño Southern Oscillation is a major ocean – atmosphere phenomenon that strongly contributes to the timing and intensity of rainfall in the tropical Pacific islands and beyond . As a consequence , it also has a major effect on the number of cases of leptospirosis . By incorporating oceanographic parameters in models , we have been able to predict leptospirosis outbreaks in New Caledonia 4 months in advance . We discuss that this forecasting delay might be used to implement timely interventions prior to the occurrence of outbreaks . Possible interventions include controlling rodent reservoir populations before population growth and maintaining sewage networks or river banks . The major impact of El Niño on the climate of Pacific Islands Countries and Territories suggests that other countries might be able to build similar predictive models .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"atmospheric",
"science",
"tropical",
"diseases",
"social",
"sciences",
"geoinformatics",
"population",
"modeling",
"neglected",
"tropical",
"diseases",
"veterinary",
"science",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"veterinary",
"diseases",
"zoonoses",
"geography",
"infectious",
"disease",
"modeling",
"environmental",
"systems",
"modeling",
"leptospirosis",
"climatology",
"earth",
"sciences",
"climate",
"modeling",
"biology",
"and",
"life",
"sciences",
"computational",
"biology"
] |
2014
|
El Niño Southern Oscillation and Leptospirosis Outbreaks in New Caledonia
|
The strengths of association mapping lie in its resolution and allelic richness , but spurious associations arising from historical relationships and selection patterns need to be accounted for in statistical analyses . Here we reanalyze one of the first generation structured association mapping studies of the Dwarf8 ( d8 ) locus with flowering time in maize using the full range of new mapping populations , statistical approaches , and haplotype maps . Because this trait was highly correlated with population structure , we found that basic structured association methods overestimate phenotypic effects in the region , while mixed model approaches perform substantially better . Combined with analysis of the maize nested association mapping population ( a multi-family crossing design ) , it is concluded that most , if not all , of the QTL effects at the general location of the d8 locus are from rare extended haplotypes that include other linked QTLs and that d8 is unlikely to be involved in controlling flowering time in maize . Previous independent studies have shown evidence for selection at the d8 locus . Based on the evidence of population bottleneck , selection patterns , and haplotype structure observed in the region , we suggest that multiple traits may be strongly correlated with population structure and that selection on these traits has influenced segregation patterns in the region . Overall , this study provides insight into how modern association and linkage mapping , combined with haplotype analysis , can produce results that are more robust .
Association mapping , which was developed as a necessity for large-scale human studies , is commonly used in conjunction with family ( linkage ) mapping in plant and animal genetic studies . The application of association mapping for plants was originally assessed in Thornsberry J . M . ( 2001 ) [1] with Buckler as senior author . It was concluded that association mapping offers higher resolution than linkage mapping due to quicker linkage disequilibrium ( LD ) decay , that structured association mapping is crucial for controlling false positives arising from population structure , and that Dwarf8 ( d8 ) ( RefGen_v2 position: Chr . 1; 266 , 094 , 769–266 , 097 , 836 bp ) is associated with flowering time . This initial study has been cited extensively , and has been the basis of several reanalyses of d8 . New data and statistical tools give us the opportunity to reevaluate this locus . Results show that the d8 associations reported by Thornsberry et al . ( 2001 ) are likely false positives ( i . e . , spurious associations ) , which resulted from insufficient correction of population structure . Indeed , the application of association mapping to animal and plant studies has been very successful , culminating in many important findings [2]–[10] . In this light , the Thornsberry et al ( 2001 ) study has attracted a lot of interest to the area and led to more studies and the development of techniques to control for population structure and familial relatedness . When the phenotype is strongly correlated with population structure ( e . g . , flowering time ) , it is often difficult to obtain statistically significant results when the models used include covariates accounting for population structure . This leads to uncertainty when determining which associated sites are causative . Thus , linkage mapping is a valuable complementary approach in these situations , and in maize , large-scale connected mapping populations issued from diverse founders have been developed [11] , [12] in order to conduct joint linkage-association analyses [10] , [13] , [14] . A major issue with association studies is false positives . In particular , indirect associations that are not causal will not be eliminated by increasing the sample size or the number of markers [15] . The main sources of such false positives are linkage between causal and noncausal sites , more than one causal site , and epistasis . These indirect associations are not randomly distributed throughout the genome and are less common than false positives arising from population structure . This makes them more difficult to control for than false positives arising from population structure . The identification of a statistically significant association between a genotypic marker and a trait is considered to be proof of linkage between the phenotype and a casual site . This assumption is true for random mating populations with fast LD decay [16] . However , it is important to consider that population structure is typically present in association panels and it has an impact on the results . Population structure exists among all species in forms such as colonies , ethnic groups , and other subdivisions based on selection or geography . Typically , population structure leads to spurious associations between markers and the trait [17] . The ability to account for population structure in a given data set is influenced by the population size , the number of markers , the level of admixture , and the divergence in allele frequency between the subpopulations [18] . One commonly used method for controlling population structure is structured association ( SA ) , which relies on randomly selected markers from the genome to estimate population structure . This estimate is then incorporated into the association analysis [16] , [18] , [19] . Another methodology for controlling population structure is to conduct a principal component analysis ( PCA ) [20] , [21] . This approach summarizes the variation observed across all markers into a smaller number of underlying component variables . One interpretation of these principal components relates them to separate , unobserved subpopulations from which the individuals in the data set originate . The loadings ( i . e . , coefficient values ) of the individuals for each principal component describe their relationship to the subpopulations . Both SA and PCA are limited to correcting for spurious associations by clustering on a global level of genetic variation . Thereby , they do not adequately capture the relatedness between individuals . Correcting for population structure is not sufficient to eliminate all false positives . Therefore , the unified mixed linear model ( MLM; also called the Q+K model ) [22] was developed to further reduce the false positive rate by controlling for both population structure and cryptic familial relatedness . This approach uses a mixed model framework that has traditionally been used by animal geneticists [23] , [24] . Specifically , covariates accounting for population structure are included as fixed effects ( Q ) , and the individuals in the association panel are included as random effects . A kinship matrix ( K ) is calculated to estimate the variance-covariance between the individuals . Typically , the covariates used in the unified MLM are either principal components of the markers or covariates from SA approaches ( e . g . , STRUCTURE [17] ) . The advantages of the MLM are that it crosses the boundary between family-based and population-based samples . However , not all associations that are eliminated will be false . If a polymorphism is perfectly correlated with population structure , it is not possible to differentiate between true and false positives . The initial study by Thornsberry et al . ( 2001 ) identified nine polymorphisms within d8 [25] that were associated with variation in flowering time in an association panel consisting of 92 diverse inbred lines . The most significant site was an 18 bp deletion ( RefGen_v2 position: Chr . 1; 266 , 094 , 529 bp ) in the promoter region . A 6 bp indel ( RefGen_v2 position: Chr . 1; 266 , 095 , 483 ) was also identified . This allele is over-represented in Northern Flint lines and is located near a Src Homology 2-like domain , which is an important binding domain within this class of transcription factors . The initial association analysis was performed using logistic regression analyses , accounting for population structure . Population structure was estimated as a modification of SA using STRUCTURE software [18] with k = 3 . Using a general linear model ( GLM ) without population structure , Andersen et al . ( 2005 ) obtained similar results for six of the nine d8 polymorphisms identified by Thornsberry et al . ( 2001 ) . However , when including population structure in the model , ( using STRUCTURE with both k = 2 and k = 3 subpopulations ) , it was found that the association results were overestimated . Each subpopulation was also analyzed separately , and a spurious association was still detectable [26] . Camus-Kulandaivelu et al . ( 2006 ) examined the association between d8 and flowering time using a panel of 375 inbred lines ( including the 92 from the initial study ) as well as a panel consisting of 275 traditional landraces from American and European origins [27] . Population structure was estimated using STRUCTURE , and association analysis was performed using both GLM and logistic regression . Their analysis revealed that the 6 bp indel at 266 , 095 , 483 bp ( identified in Thornsberry et al . , 2001 ) was spuriously associated with flowering time when covariates accounting for population structure were not included . In contrast , no association between d8 and flowering time was detected in the inbred panel when accounting for population structure . However , this spurious association was still detectable in the traditional landraces panel , including Andean material that has no relationship to the Northern Flint material . The d8 gene produces a signaling factor involved in the gibberellin pathway . Gibberellins are types of endogenous plant growth regulators [28] . Maize d8 and wheat Rht-B1/Rht-D1 have been shown to be orthologous of the GAI gene [25] . Mutants of d8 have severe height phenotypes due to alterations of the DELLA domain . In maize , these are dominant , gain-of-function mutations , suggesting that d8 is a negative regulator . Conversely , recessive mutants of the GAI gene in Arabidopsis result in loss-of-function , specifically in polypeptides truncated upstream of the SH2-like domain . As a consequence , the gene product does not function as a negative regulator , resulting in normal height phenotypes [1] . Two evolutionary processes have likely impacted the d8 locus . First , the associated allele , specifically the 6 bp indel reported in Thornsberry et al . ( 2001 ) , is related with Northern Flint maize . Maize originated from southern Mexico , where there are long growing seasons and high temperatures . As maize agriculture expanded from Mexico through the Southwestern United States to the Eastern United States ( with its shorter growing season and lower temperatures ) , a severe bottleneck occurred in maize diversity , resulting in the Northern Flint subpopulation [29] . The bottleneck created extensive long range LD in this subpopulation . Northern Flints were substantially isolated from all other maize subpopulations [29] until the introduction of the Southern Dents in the 1600s [30] . Additionally , the d8 locus is located only 347 , 057 bp from the tb1 ( teosinte branched1 ) locus ( RefGen_v2 position: Chr . 1; 265 , 745 , 979–265 , 747 , 712 bp ) , which is one of the key genes involved in maize domestication [31] . The tb1 locus lost much of its diversity during the domestication process [31] , [32] . The original d8 study [1] identified evidence of purifying selection with substantial diversity loss; however , there was little LD identified in the region between d8 and tb1 . Although unconfirmed , some Northern Flint allied germplasm ( e . g . sweet corn , P39 ) have a morphology that looks like the undomesticated tb1 phenotype . It is likely that the region around d8 and tb1 has been through a bottleneck with multiple selective sweeps , resulting in complex extended haplotypes . Most of the loci controlling flowering time in maize have been identified through QTL studies . Of these , only d8 and vegetative to generative transition 1 ( vgt1 ) have been confirmed with association and fine mapping [33] . Located on chromosome 8 , vgt1 is arguably the most important flowering time locus in maize . It contains an APETALA2-like gene , ZmRap2 . 7 , which is controlled by an enhancer region about 70 kb upstream [33] . The association between vgt1 and flowering time is supported by a study conducted in the maize nested association mapping ( NAM ) population , where a major QTL was identified in this region [11] . This study also detected an allelic series at this QTL , suggesting that more than one causative allele is present . One of these alleles is from northern germplasm and is in linkage with a MITE whose association with early flowering time was confirmed in the NAM population [11] . Although the lack of the vgt1 early flowering allele did not completely explain the late flowering time , a SNP identified in the ZmRap2 . 7 gene showed association with the late flowering effect [11] . An association study by Ducrocq et al . ( 2008 ) [34] reported P-values for vgt1 association several magnitudes lower that those obtained by Salvi et al . ( 2007 ) [33] . Both studies accounted for population structure . Compared to Salvi et al . ( 2007 ) [33] , Ducrocq et al . ( 2008 ) used a more genetically diverse and larger association panel , including a higher number of lines from Northern Flint and European germplasm [34] . In the case of d8 , the association between the site and the trait becomes less significant , and even undetectable , when increasing the number of lines examined . This supports no association between the 6 bp indel in d8 and flowering time in maize . Including d8 in the model when performing association mapping for flowering time does not change the result for the SNPs in vgt1 [34] . This indicates that there is no interaction between the two loci . The purpose of this study was to reanalyze the work of Thornsberry et al . ( 2001 ) utilizing some of the latest association mapping methodologies and data sets . This study compared association results from various statistical approaches using a maize diversity panel and the NAM population [11] , [12] . Single nucleotide polymorphisms ( SNPs ) and insertions/deletions ( indels ) from recent genotyping efforts ( e . g . , HapMap sequencing from Gore et al . 2009 [35] and Chia et al . 2012 [36] ) were used to evaluate these various approaches and the d8 association .
The results from the Thornsberry et al . ( 2001 ) study showed significant association at both the 18 bp deletion ( 266 , 094 , 829 bp ) in the promoter region and the 6 bp indel ( 266 , 095 , 483 bp ) . Our reanalysis of the two sites using the Q model and a significantly larger association panel ( consisting of 282 lines ) resulted in less significant associations at both loci ( Table 1 ) . By increasing the number of lines we are able to obtain a larger sample size within each of the subpopulations and thus , more accurately estimate the underlying population structure ( i . e . , Q ) . Sampling has a larger effect on some sites than others . The 6 bp indel is more significantly associated with flowering time in the smaller population ( 92 lines ) than it is in the 282 association panel analyzed with MLM ( K model ) without controlling for population structure , but controlling for familial relatedness . The site is , in fact , carried by Northern Flint lines , which are underrepresented in the smaller population . The results for the 282 association panel suggest that the GLM ( Q ) approach overestimates the association . In contrast , the MLM ( Q+K ) approach , which accounts for both population structure and relatedness between individuals , gives a moderately significant association between the 6 bp indel ( P-value = 0 . 0127 ) and flowering time variation ( Table 1 ) . The proportion of the genetic variation explained by the different models varies significantly . In this study , the best models are the Q+K and K models ( the latter being a MLM that only includes familial relatedness between individuals as random effects ) because they explain the highest amount of the genetic variance ( Table 2 ) . The reason for the minimal difference between the two models is that K most likely controls for the majority of the relatedness between individuals . This study confirms the weak association between the 6 bp indel in d8 and flowering time analyzed using both GLM and MLM approaches ( Table 1 ) . However , the association is not as significant as previously reported by Thornsberry et al . in 2001 . Additionally , the GLM and MLM analyses of the 282 association panel imply there is no association between the 18 bp deletion in d8 and flowering time ( Table 1 ) . The initial study by Thornsberry et al . ( 2001 ) found this site to be the most significant . Our results from the Q+K and K models yielded a more significant P-value for the 92 association panel than the 282 association panel . We also sequenced a 3 bp indel ( 266 , 097 , 198 bp ) , which is present in tropical late-flowering lines when we examined sequences available at NCBI . However , new genotypic data for the 282 association panel suggest that there is no association between this site and variation in flowering time in maize ( Table 1 ) . Our study confirms the results presented by Camus-Kulandaivelu et al . ( 2008 ) [37] , that there are regions between d8 and tb1 associated with variation in flowering time ( Table 1 ) ( Figures S3 and S4 ) . However , these sites are moderately significant at α = 0 . 05 when using the K and Q+K models . Association mapping of d8 on other traits results in a number of weak associations with other traits , in addition to flowering time ( e . g . , plant height , ear height , and node number ) ( Table 3 ) . All the associations are in the same range of significance as flowering time . No clear pattern can be observed between correlation among traits except for what can be expected ( e . g . , the high correlation between days to silk and days to anthesis ) ( Figure 1 ) . Collectively , these results undermine the conclusion that d8 is of more importance for flowering time than any of the other traits . From a genome-wide perspective , there were a large number of sites with a similar degree of association ( from the MLM approach ) with flowering time as d8 ( Figure 2A ) . The contrasting results from the various models fitted at the SNPs in the genomic regions surrounding d8 and tb1 are illustrated in Figure 2B and 2C . In particular , the GLM model overestimated the significance of the results in comparison to the Q+K and K models . GWAS of flowering time detected SNPs within d8 that have a weak statistically significant association at α = 0 . 05 ( Figure 2C ) . Linkage mapping of flowering time in the NAM population detected a number of QTL . A small QTL ( P-value = 0 . 0127 ) colocalized with d8 ( RefGen_v2 position: Chr . 1; 269 , 321 , 476–269 , 322 , 794 bp ) , supporting the association identified by association mapping ( Figure 3 ) . In the initial study by Thornsberry et al . ( 2001 ) , the effect of d8 was estimated to be between 7–10 days . The d8 polymorphism should be in three of the mapping families , and modest effects are seen in the right direction for all three , but the estimated effect is always less than half a day . Additionally , many of the subfamilies appear to have other QTL along this section of chromosome 1 ( RefGen_v2 position: Chr . 1; 231 , 701 , 106–231 , 703 , 173 bp and Chr . 1; 286 , 977 , 415–287 , 063 , 457 bp ) , but the favored positions are millions of base pairs away . It is quite possible that the mapping position of these joint linkage QTL could be synthetic , but there is little to no support for a QTL in this exact region . A GWAS in the NAM population for flowering time using 26 . 5 million segregating SNPs was performed [35] , [36] . This approach in the NAM population offers in-depth power and resolution because it utilizes both historic and recent recombination . No significant sites were identified in the region of d8 ( Figure S5 ) . This supports the result that d8 is not associated with flowering time . Hapmap data [35] , [36] suggest extended haplotypes for Northern Flint lines in the region of d8 . Data show modest Fst between temperate and tropical subpopulations . However , there could potentially be differences in diversity between these two groups and Northern Flint lines . Hapmap data are only available for a few Northern Flint lines , which limits these studies . GBS SNPs were used to examine the range of LD decay within the different subpopulations ( Northern Flint , stiff stalk , non-stiff stalk , and tropical ) of the 282 association panel . Extended LD is observed for the Northern Flint lines compared to the other subpopulations . Likewise , the stiff stalk lines , which were only founded from 16 inbred lines , also show a pattern of extended haplotypes , although not as extreme as the Northern Flints ( Figure 4 ) . The extended haplotype pattern in the Northern Flints make it difficult to control for false positives and to identify the causative SNP using association mapping . The 6 bp indel in d8 is carried by Northern Flint lines . When we examine the LD between the 6 bp indel and the 13 , 815 high coverage GBS SNPs on chromosome 1 ( Figure 5 ) , an extended area around the 6 bp exhibits fairly high values of R2 . This is additional evidence that extended haplotypes exist in the Northern Flint lines in the d8 region . In fact , there are two regions with high LD at 20 Mbp and 0 . 9 Mbp away , which contain previously identified domestication gene candidates ( i . e . , GRMZM2G034217 RefGen_v2 position: chr . 1 246 , 720 , 001–247 , 030 , 000 , a mitochondrial transcription termination factor ) [38] . In contrast , LD between the MITE in vgt1 and the 7 , 539 GBS SNPs on chromosome 8 ( Figure S6 ) show sites with high R2 values close to the position of the MITE , but LD decays much more rapidly . To test for two-way interaction between the 6 bp and 18 bp indels and the MITE , a series of mixed models including two-way interaction terms were fitted . The most significant interaction was between the 18 bp indel and the MITE ( P-value 0 . 0418 ) . However , this association is not likely to be statistically significant after controlling for the multiple testing problem across the entire genome .
The results underscore the importance of properly accounting for population structure in association studies . The analysis in Thornsberry et al . ( 2001 ) divided their 92 association panel into three subpopulations . This subdivision did not fully account for population structure , and thus , the effect of the d8 allele carried by Northern Flint lines was overestimated . In contrast , our study accounted for population structure using both k = 3 ( stiff stalk , non-stiff stalk , and tropical ) and k = 5 ( stiff stalk , non-stiff stalk , tropical , sweet corn , and popcorn ) subpopulations . This was important for sites such as the 6 bp indel within d8 , which is present at a higher frequency in Northern Flint lines ( which includes sweet corn ) , and this signature of population structure was unaccounted for when k = 3 was used . Of all the models tested , the Q+K model [22] was the most suitable approach for analyzing the 282 association panel because it controls for both population structure and cryptic familial relatedness . It is especially important to control for the latter with traits such as flowering time , which is highly correlated with population structure . Additionally , the Q+K model is beneficial for association studies because Q and K capture different types of long range LD [22] , [39] . In general mixed models sufficiently account for population structure and familial relatedness . In contrast , false positives arising from other sources , although rare , are typically unaccounted for in association studies . For example , spurious associations could arise from markers that are in long-range LD with causative polymorphisms . Additionally , causative polymorphisms for one trait may not necessarily be causal for another highly correlated trait ( and , hence a spurious association ) , but will be statistically associated with both traits . Finally , when a trait is controlled by multiple loci in LD , it is likely that the site with the largest effect is an indirect association . One reason for this result arises from differing minor allele frequencies among the causal sites . All three of these types of false positives do not occur randomly across the genome and thus , they are more challenging to eliminate . Haplotype-based association studies is one approach for addressing many of these issues . Nevertheless , multiple sites , selection for multiple traits , and population structure result in spurious associations and these need to be accounted for when performing association studies . Association mapping is limited when the trait analyzed is correlated with population structure . However , linkage mapping can overcome this problem by crossing individuals with known relatedness , spurious associations can be broken . In this study , we were able to detect a small QTL at the general location of d8 . However the favored QTL locations are on both sides of d8: RefGen_v2 position Chr . 1; 231 , 703 , 173–287 , 063 , 457 . Association results suggest that the majority of the QTL effects detected around d8 are from rare extended haplotypes that include other linked QTLs . Another possible explanation for the weakness of the QTLs detected is the population sampled . The associated haplotype is present only in Northern Flint lines , which are underrepresented in the population . Flowering time is strongly correlated with population structure . Our study showed that d8 had a very small effect on flowering time . One possible explanation for this result is that d8 is associated with another trait that was selected along with flowering time ( e . g . , cold tolerance ) . This hypothesis is supported by the extended haplotype pattern observed in Northern Flint lines as well as the associations that are detected with traits like plant height and node number . Northern Flint lines are underrepresented in both the 282 association panel and the NAM population , and it is difficult to scrutinize associations with low allele frequencies . Northern Flint lines have been shown to be distinctive compared to other subpopulations , especially in regions like d8 and tb1 , which have been under selection pressure . Consistent with the findings of previous studies on the d8 locus , we observed a strong correlation between d8 and population structure . Thus , the functional site of d8 is not likely to be involved in flowering time . Indeed , d8 and tb1 are strong integrators in plant signals that are adjacent to each other on chromosome 1 . However , our results demonstrated that these signals are a signature of population structure instead of true biological signals . The extended LD in the Northern Flint lines around d8 supports the hypothesis that this gene is regulated in a similar manner as tb1 and vgt1 . For both tb1 and vgt1 , cis-acting regulatory sites located more than 50 kb from the actual genes have been shown to be the functional regions and not the genes themselves [31] , [33] , [40] . Signatures of selection on d8 have been observed in teosinte [41] . Because apical dominance ( tb1 ) and gibberellin signaling ( d8 ) have both played key roles for domestication phenotypes , it is likely that the genomic region surrounding d8 and tb1 has been under selection since early maize domestication . Northern Flint lines differ from Corn Belt dent lines in a number of traits such as leaf angle , plant height , and cold tolerance . Thus , the long range LD block around d8 could be a signature of selection from the development of Northern Flint lines that happens to be associated with one of these traits distinguishing Northern Flint from Corn Belt dent . Consequently , it has been possible to detect a weak association between flowering time and d8 because of the correlation between flowering time and the Northern Flint specific traits due to population structure . Using this rationale , it may be possible to detect associations between the d8 locus and phenotypes such as carbon allocation and harvest index , when considering the differences in the usage of Northern Flints ( sweet corn and silage ) and Corn Belt dent . The basic d8 associations identified in Thornsberry et al . ( 2001 ) have been replicated by other independent groups [26] , [27] , but population structure has always remained a consistent issue . This reanalysis using SNPs within and near d8 suggests that these associations are either incorrect or vastly overestimated . This work implemented more powerful statistical approaches , germplasm resources , and whole genome sequencing data , enabling a more thorough understanding of this locus . This analysis underscores the importance of controlling for population structure . All three previously published studies on the d8 locus illustrate how naïve association results overestimated effect sizes . In our study , we used the unified MLM to control for both population structure and relatedness between individuals , which are more accurate in effect estimation and give a truer level of significance . Even in species with rapid LD decay , like maize , it is possible to have subpopulations that can exhibit LD many orders of magnitude greater than the average length . This long range LD resulted in the extended haplotype lengths observed in Northern Flint lines for the genomic region surrounding d8 . Northern Flint lines are underrepresented in the association panel , which makes it difficult to accurately account for the population structure of this subpopulation . Another issue is the strong correlations between traits . It is very likely that in the case of d8 there has been selection for other correlated traits , such as cold tolerance . Because of the correlation between the population structure and flowering time , we can detect a weak association between flowering time and d8 , but d8 does not actually have an effect on time of flowering . Genes like d8 have been targets of strong selection and , as such , are among the hardest to identify in GWAS and accurately estimate their effect size . NAM-like linkage populations with bi-parental crosses in a reference design to minimize population structure may be necessary for dissecting the most structured traits . Although our results strongly suggest that the previously reported association between d8 and flowering time is an artifact of population structure , further research on this complex locus is warranted . The long range LD present at d8 for Northern Flint lines is a signature of selection , and it is important to determine the traits that are regulated by this gene . By applying the appropriate statistical models , we have shown that flowering time is not one of these traits .
The association panel consists of 282 diverse maize lines that have been previously described [42] . These lines can be subdivided into five major subpopulations , namely stiff stalk , non-stiff stalk , tropical or semitropical lines ( related to the non-stiff stalk lines ) , sweet corn and popcorn . The association panel includes the 25 founder lines of the NAM population . The maize NAM population consists of 5 , 000 RILs ( Recombinant Inbred Lines ) derived from crossing B73 with 25 diverse maize inbred lines , and then selfing for 5 generations [43] . Phenotypic data were collected from the NAM population and the 282 association panel , grown in eight environments including Ithaca , NY , Clayton , NC , Champaign , IL , and Colombia , MO , during the summers of 2006 and 2007 . Flowering time was measured separately for female flowers ( number of days-to-silk ) and male flowers ( days-to-anthesis ) from the day of planting . The flowering date was defined as the day when the anthers or silk were visible on 50% of all plants within a row . DNA sequence data were obtained for d8 from Thornsberry et al . ( 2001 ) , available at NCBI . Primers were designed for PCR amplification of gene fragments of interest from the 282 lines in the association panel . Each PCR product was cleaned by treating the samples with Exonuclease ( ExoI ) and Shrimp Alkaline Phosphatase ( SAP ) and incubated at 37°C for 3 min followed by 80°C for 10 min . The samples were prepared for sequencing using a mixture with a total volume of 10 µl containing 0 . 7 µl forward primer , 0 . 7 µl reverse primer ( 5 pmol/µl ) , 0 . 5 µl Big Dye terminator , 1 . 7 µl 5× sequencing buffer , 7 . 1 µl distilled water and the PCR product . The thermal cycler was set on the following program: Initial denaturation at 96°C for 4 min , followed by 30 cycles at 96°C for 10 sec , 50°C for 5 sec and 60°C 4 min , with a final , incubation at 10°C . Sanger ( 3730XL ) DNA sequencing was performed using an Applied Biosystems Automated 3730 DNA Analyzer . The software BioLign alignment and multiple contig editor with codon code phred-phrap analysis was used for alignment using consensus sequence contigs and sequence quality scores . The alignments from NCBI were also used to reanalyze the results published in the initial study by Thornsberry et al . ( 2001 ) . To obtain sequence data for the region between d8 and tb1 , the same protocol was used as described above . However , primer sequences were obtained from Camus-Kulandaivelu et al . ( 2008 )
|
Eleven years ago , association mapping was a cutting-edge tool used to identify regions of a genome associated with phenotypic variation . One of the first association studies performed in plants was reported in Thornsberry , et al . ( 2001 ) . Since then , researchers continued to develop new and improved genotyping , phenotyping , and statistical methods to examine the relationship between genotype and phenotype . Reanalysis of the old data for the d8 locus and flowering time , as well as new and improved data sets , gives us a unique opportunity to examine the strengths and weaknesses of association studies . These new analyses reveal that the results reported in 2001 significantly overestimated the association between genotype and phenotype , in particular the estimated effect size . The key issues with the Thornsberry et al . ( 2001 ) study were lack of control for population structure and relatedness between individuals , as well as a potential confounding between the phenotype and the population structure examined . The new analysis demonstrates a marginal association between d8 and flowering time , and a minimal effect ( if any ) .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"mathematics",
"plant",
"biology",
"crops",
"genetics",
"statistics",
"biology",
"genomics",
"genetics",
"and",
"genomics",
"agriculture"
] |
2013
|
Lessons from Dwarf8 on the Strengths and Weaknesses of Structured Association Mapping
|
The signaling of Toll-like receptors ( TLRs ) is the host's first line of defense against microbial invasion . The mitochondrion is emerging as a critical platform for antiviral signal transduction . The regulatory role of mitochondria for TLR signaling remains to be explored . Here , we show that the mitochondrial outer-membrane protein MARCH5 positively regulates TLR7 signaling . Ectopic expression or knockdown of MARCH5 enhances or impairs NF-κB-mediated gene expression , respectively . MARCH5 interacts specifically with TANK , and this interaction is enhanced by R837 stimulation . MARCH5 catalyzes the K63-linked poly-ubiquitination of TANK on its Lysines 229 , 233 , 280 , 302 and 306 , thus impairing the ability of TANK to inhibit TRAF6 . Mislocalization of MARCH5 abolishes its action on TANK , revealing the critical role of mitochondria in modulating innate immunity . Arguably , this represents the first study linking mitochondria to TLR signaling .
Germline-encoded pattern recognition receptors ( PRRs ) play pivotal roles in sensing a wide range of invading pathogens , via recognizing conserved microbial signature molecules ( PAMPs , pathogen associated molecular patterns ) . As molecular switches that register microbial infection , these receptors promptly initiate innate immune responses and subsequently prime the adaptive immune system to eliminate the pathogens [1] , [2] , [3] . Toll-like receptors ( TLRs ) are one major class of PRRs . To date , 13 members of the TLR family have been identified in mammals , of which TLR4 recognizes lipopolysaccharide ( LPS ) from gram-negative bacteria , and TLR7 senses viral single strand RNA ( ssRNA ) [2] . Both TLR4 and TLR7 can induce the robust expression of proinflammatory cytokines and type I interferons ( IFNs ) . Recent years have witnessed an incredible gain in knowledge about the TLR signaling cascades . For example , in response to stimuli , TLR7 triggers the recruitment of MyD88 via its Toll-interleukin 1 receptor ( IL-1R ) homology ( TIR ) domain , which in turn recruits IRAK4 and IRAK1 . IRAK4 then activates IRAK-1 by phosphorylation . As a result , the IRAKs dissociate from MyD88 and interact with TRAF6 , a ubiquitin E3 ligase [4] , [5] , [6] . Together with Ubc13 and Uev1A , TRAF6 catalyzes the formation of lysine 63 ( K63 ) -linked poly-ubiquitin chains , which serve as the anchoring platform for a protein complex that includes TRAF6 , TRAF3 , IKKα , and IRF7 , leading ultimately to the induction of type I IFNs and ISGs ( interferon-inducible genes ) [7] , [8] . In the meantime , TAK1 is recruited to the TRAF6 protein complex , resulting in the activation of NF-κB and the induction of proinflammatory cytokines [9] . Given the critical roles of TLR signaling in innate immunity , multiple layers of stringent regulations are employed to ensure that the strength and duration of the TLR signal is appropriate for any given immune response . Several proteins have been demonstrated to modulate the TLR signaling pathways , such as A20 , CYLD , IRAKM , ST2 , SIGIRR and SOCS1 [10] , [11] , [12] , [13] , [14] , [15] , [16] . Importantly , TANK ( also called I-TRAF ) was initially characterized as a TRAF binding protein [17] , [18] . Recently , the in vivo function of TANK has been further clarified . Surprisingly , TANK is not essential for interferon induction and instead is a potent negative regulator for TLR-mediated induction of proinflammatory cytokines [19] . How TANK specifically modulates NF-κB signaling upon TLR activation remains to be determined . Mitochondria are rapidly emerging as important platforms for intracellular antiviral signaling . MAVS ( also known as IPS-1/VISA/Cardif ) is the first mitochondrial protein identified as a critical component of the RIG-I/MDA5 signaling pathway [20] . Following this , several more mitochondrial proteins have been implicated in modulating this same antiviral signaling pathway , such as NLRX1 , STING ( also known as MITA ) and MFN1 [21] , [22] , [23] , [24] . We have recently discovered that the mitochondrial outer-membrane receptor TOM70 mediates IRF3 activation downstream of MAVS [25] , [26] , [27] . Notably , whether any mitochondrial protein ( s ) can regulate TLR signaling remains an open question . It was recently reported that two ubiquitin E3 ligases constitutively express on mitochondria , MARCH5/MITOL/RNF153 and GIDE ( Growth Inhibition and Death E3 Ligase ) . MARCH family proteins are characterized by harboring a RING-CH domain and multiple trans-membrane domains . Interestingly , many RING proteins function as E3 ubiquitin ligases . Some of the MARCH proteins ( MARCH1 , MARCH8 ) appear to either directly or indirectly modulate immune functions by controlling the surface turnover of immune regulatory molecules on the plasma membrane [28] . MARCH5 ( also named MARCH-V/MITOL/RNF153 ) was recently identified as a new member in the MARCH family [29] , [30] . Preliminary characterization uncovered MARCH5 as a novel mitochondrial protein . Until recently , little was known about the potential function of MARCH5 . Several recent studies have determined that MARCH5 can potentially modulate mitochondrial fission and the morphology of mitochondria [29] , [30] , [31] . In this study , we reveal that the mitochondrial protein MARCH5 is a novel E3 ubiquitin ligase and a positive regulator for TLR7 signaling . MARCH5 catalyzes the K63-linked poly-ubiquitination of TANK on its Lysines 229 , 233 , 280 , 302 and 306 . This modification releases the inhibitory effects of TANK toward TRAF6 . Consequently , ectopic expression or knockdown of MARCH5 enhances or impairs NF-κB-mediated proinflammatory gene expression , respectively , in response to TLR7 activation . Interestingly , the localization of MARCH5 on mitochondria is indispensable for its regulatory action , demonstrating a new role for mitochondria in the proinflammatory response . To our knowledge , this is the first report that links mitochondria to TLR signaling .
Poly-ubiquitination is a major mechanism for regulating innate immunity . Because the mitochondrial protein MAVS was recently reported to be modified by both K48- and K63- linked poly-ubiquitin [32] , [33] , [34] , we speculated whether these post-translational modifications were mediated by any mitochondrial ubiquitin E3 ligase . We noticed that MARCH5/MITOL/RNF153 is constitutively expressed in mitochondria and contains a RING domain , a signature of ubiquitin E3 ligases . To explore the potential role of MARCH5 , we screened effective siRNAs against MARCH5 ( MARCH5 siRNA 369 and MARCH5 siRNA 582 for mouse , hMARCH5 siRNA for human ) ( Fig . 1A , B ) . It was found that knockdown of endogenous MARCH5 did not affect κB-luciferase and PRDIII-I-luciferase reporters ( responsive to NF-κB and IRF3/7 activation , respectively ) stimulated by Sendai virus or transfected poly ( I:C ) . Consistently , silencing of MARCH5 produced no effect on RIG-I- or MAVS-mediated activation of κB-luciferase or IFN-β-luciferase reporters ( Fig . 1 E , F ) . Nor did MARCH5 produce any effect on MAVS poly-ubiquitination ( Fig . S7 ) . These indicate that MARCH5 is dispensable for RIG-I/MDA5 signaling . Surprisingly , activation of κB-luciferase reporter was apparently impaired upon MARCH5 depletion in response to TLR3 , TLR7 or TLR4 signaling in Raw264 . 7 cells when stimulated with extracellular poly ( I:C ) , imiquimod ( R837 ) , Loxoribine or LPS , respectively ( Fig . 1C ) . Notably , the impairment was more significant for TLR7 and TLR4 than for TLR3 . Interestingly , activation of the PRDIII-I-luciferase reporter was barely modulated by MARCH5 under these stimuli ( Fig . 1D ) . Apparently , MARCH5 could modulate TRIF- , MyD88- or IRAK1-induced κB-luciferase activity ( Fig . 1E ) . These data suggest that MARCH5 is possibly critical for NF-κB activation downstream of TLRs but is dispensable for IRF3/7 activation . In addition , confocal microscopy and fractionation analysis confirmed that MARCH5 was predominantly expressed on mitochondria ( Fig . 1G , H ) . R837 and LPS stimulation did not influence the expression of MARCH5 ( Fig . 1I , J ) . To substantiate these observations , the introduction of wild-type ( WT ) MARCH5 into Raw264 . 7 cells potentiated the activation of the κB-luciferase reporter upon R837 or LPS stimulation ( Fig . 2A ) . In contrast , the activation of PRDIII-I-luciferase reporter was unaffected by ectopic expression of MARCH5 under the same stimuli ( Fig . 2B ) . Furthermore , we investigated whether MARCH5 modulated the expression of NF-κB-responsive genes ( IL6 and TNFα ) induced by R837 using the assays of qPCR ( quantitative PCR ) and ELISA ( enzyme-linked immunosorbent assay ) . As shown in Fig . 2C and Fig . S1A , MARCH5 displayed synergic effects on the induction of NF-κB-responsive genes . Similarly , MARCH5 synergized NF-κB activation stimulated by LPS ( Fig . 2D and Fig . S1B ) . However , this did not apply to the induction of ISG15 mRNA , which is regulated by IRF3/7 ( Fig . 2C , D ) . To make the experiments more physiologically relevant , we examined whether MARCH5 regulated NF-κB-induced gene expression in primary cells . Consistently , ectopic expression of MARCH5 markedly potentiated the expression of endogenous NF-κB-responsive genes ( IL6 and TNFα ) induced by R837 in BMDCs ( bone marrow-derived dendritic cells ) ( Fig . 2E ) . To explore the potential role of MARCH5 RING domain , we generated MARCH5 C2A ( C65/68S ) and MARCH5 ΔRING ( Δ12-70 a . a . ) , both of which abrogated the potential E3 ubiquitin ligase activity of MARCH5 ( Fig . S5A ) . As shown in Fig . 2A , exogenous expression of MARCH5 C2A and MARCH5 ΔRING failed to synergize the induction of NF-κB-responsive reporter upon R837 or LPS stimulation . Consistently , the expression of NF-κB-responsive genes ( IL6 and TNFα ) induced by R837 or LPS was barely affected when MARCH5 C2A or MARCH5 ΔRING was ectopically expressed ( Fig . 2C , D , E and Fig . S1A , B ) . Apparently , ectopic expression of MARCH5 has no effect on the cell cycle or apoptosis ( Fig . S2 ) . Taken together , these data suggest that MARCH5 specifically potentiates TLR signaling , and this effect is dependent on its E3 ubiquitin ligase activity . Next , we explored the effect of MARCH5 knockdown on the expression of endogenous NF-κB-responsive genes , stimulated by R837 or LPS in Raw264 . 7 cells . As expected , MARCH5 knockdown attenuated the induction of NF-κB-responsive genes ( IL6 and TNFα ) , but not that of an IRF3/7-responsive gene ( ISG15 ) ( Fig . 3A , B and Fig . S1C , D ) . We further investigated the function of MARCH5 in primary cells . BMDCs were transfected with siRNAs against MARCH5 , followed by R837 or LPS stimulation . Consistently , knockdown of MARCH5 markedly attenuated the expression of endogenous NF-κB-responsive genes in BMDCs ( Fig . 3C , D ) . To rule out potential off-target effects of the MARCH5 siRNA , we generated several RNA interference ( RNAi ) -resistant HA-MARCH5 constructs , namely rMARCH5 WT , rMARCH5 C2A and rMARCH5 ΔRING , in which silent mutations were introduced into the sequence targeted by the siRNA without changing the amino acid sequence of the corresponding proteins . Raw264 . 7 cells were first transfected with control or MARCH5 siRNA followed by transfection of control or indicated rMARCH5 plasmids , respectively . Then the induction of IL6 mRNA was measured after R837 or LPS stimulation . As shown in Fig . 3E and Fig . 3F , the induction of IL6 was restored by rMARCH5 WT , but not rescued by rMARCH5 C2A and rMARCH5 ΔRING . Apparently , knockdown of MARCH5 influences neither the cell cycle nor apoptosis ( Fig . S3 ) . Collectively , these results indicate that MARCH5 is a positive regulator of TLR-mediated NF-κB activation . To address the mechanism of MARCH5 action , we explored whether MARCH5 could interact with TLR signaling proteins . A co-immunoprecipitation assay revealed that MARCH5 interacted with TRAF6 , TANK and MAVS but did not interact with TRAF3 , NEMO or RIG-I ( Fig . 4A ) . Interestingly , MARCH5 did not catalyze the ubiquitination of any of the indicated proteins except TANK ( Fig . S7 ) . Therefore , we further probed the interaction between MARCH5 and TANK . As shown in Fig . 4B , HA-MARCH5 co-immunoprecipitated with Flag-TANK , but not with control IgG . Similarly , Flag-TANK co-immunoprecipitated with HA-MARCH5 , but not with control IgG ( Fig . 4C ) . Notably , HA-MARCH5 C2A and HA-MARCH5 ΔRING could interact with Flag-TANK as well ( Fig . 4B , C ) . We subsequently confirmed the endogenous interaction between MARCH5 and TANK ( Fig . 4D ) . Interestingly , the endogenous interaction between MARCH5 and TANK was enhanced by R837 stimulation ( Fig . 4D ) . A series of Flag-TANK deletion mutants were generated and individually transfected into HEK293T cells along with HA-MARCH5 . It was observed that the C terminal region of TANK ( amino acids 190 to 413 ) mediated this interaction ( Fig . 4E ) . We went on to investigate the sub-cellular localization of endogenous TANK . Confocal microscopy revealed that TANK displayed a punctate staining pattern in the cytoplasm of resting cells . Interestingly , TANK partially co-localized with mitochondria upon R837 or LPS stimulation , suggesting that TANK was dynamically recruited to mitochondria in response to these stimuli ( Fig . S4 ) . Collectively , these results indicate that MARCH5 is a new TANK binding protein in vivo . A couple of recent studies implicated MARCH5 as an E3 ubiquitin ligase that catalyzes the ubiquitination of hFis1 and Drp1 , causing their degradation [31] . In vitro ubiquitination assays confirmed that the MARCH5 RING domain can catalyze the formation of both K48- and K63-linked polyUb chains , whereas MARCH5-C2A or MARCH5 ΔRING cannot ( Fig . S5 ) . Our above data revealed the importance of this E3 ubiquitin ligase activity for regulating TLR signaling ( Fig . 2 ) . Therefore , we wondered whether TANK was a new ubiquitination target of MARCH5 . In HEK293T cells , Flag-TANK and His-Ubiquitin was co-transfected with MARCH5 ( WT ) , MARCH5 C2A , MARCH5 ΔRING or GIDE , respectively . The cell lysates were subjected to immunoprecipitation of Flag-TANK ( Fig . 5A ) or Ni-NTA pulldown of His-Ubiquitin ( Fig . 5B ) . Then the precipitates were probed with indicated antibodies as seen in Fig . 5A and Fig . 5B . Notably , TANK was markedly poly-ubiquitinated in the presence of MARCH5 ( WT ) . In contrast , MARCH5 C2A , MARCH5 ΔRING and GIDE could not catalyze ubiquitination of TANK ( Fig . 5A , B ) . In addition , ectopic-expression of MARCH5 enhanced poly-ubiquitination of TANK in Raw264 . 7 cells upon R837 stimulation , whereas MARCH5 C2A had no such effect ( Fig . 5C ) . Consistently , knockdown of MARCH5 attenuated this poly-ubiquitination of endogenous TANK ( Fig . 5D ) . These data indicate that MARCH5 is a novel E3 ubiquitin ligase for TANK . It was recently reported that TANK was modified by K63-linked polyubiquitination via ( an ) unknown E3 ubiquitin ligase ( s ) [35] . We further tested whether MARCH5 could fulfill this function . A panel of ubiquitin mutants was employed including those containing a point mutation at lysine 48 ( K48R ) or 63 ( K63R ) or lacking all lysines except K48 ( K48-only ubiquitin ) or K63 ( K63-only ubiquitin ) ( Fig . 5E upper panel ) . As expected , MARCH5 catalyzed TANK poly-ubiquitination in the presence of wild type ubiquitin . Importantly , TANK was poly-ubiquitinated as well when using the K48R or K63 ubiquitin whereas poly-ubiquitination of TANK disappeared when using the K63R or K48 ubiquitin ( Fig . 5E lower panel ) . This firmly established that MARCH5 facilitated the synthesis of K63- rather than K48- linked poly-ubiquitin chains onto TANK . To identify the potential ubiquitination sites on TANK , we carried out a systematic lysine ( K ) to arginine ( R ) mutation scanning . It was observed that ubiquitination of the TANK ( 5M ) mutant was markedly impaired ( Fig . 5F ) . Notably , the TANK ( 5M ) mutant could interact with MARCH5 as well as wild-type TANK ( Fig . S6 ) , indicating that these five lysines , 229 , 233 , 280 , 302 and 306 , were the major ubiquitination sites on TANK . Consistently , ectopic expression or knockdown of MARCH5 apparently did not affect the stability of endogenous TANK ( Fig . 5G , H ) . Recently , TANK was found to negatively regulate TLR-mediated induction of pro-inflammatory cytokines [19] . We confirmed this observation in Raw264 . 7 cells , i . e . , knockdown of TANK resulted in augmented cytokine production ( IL6 and TNFα ) in response to R837 stimulation ( Fig . 6A ) . As TRAF6 is auto-ubiquitinated in response to TLR stimuli , we tested whether TANK influenced the auto-ubiquitination of TRAF6 . Interestingly , R837-induced TRAF6 auto-ubiquitination was enhanced upon endogenous TANK depletion ( Fig . 6B ) . This indicated that TANK inhibited NF-κB activation by suppressing TRAF6 auto-ubiquitination . We went on to address whether TANK ubiquitination could release its inhibitory effects toward TRAF6 . Raw264 . 7 cells were transfected with control or MARCH5 plasmids followed by R837 stimulation . Intriguingly , TRAF6 auto-ubiquitination was enhanced in the presence of the wild type MARCH5 ( Fig . 6C ) . In contrast , the ubiquitination of TRAF6 was not affected when MARCH5 C2A was introduced ( Fig . 6C ) . Reciprocally , knockdown of MARCH5 inhibited the ubiquitination of TRAF6 ( Fig . 6D ) . Notably , suppression of TRAF6 ubiquitination caused by MARCH5 knockdown could be reversed by simultaneously knocking down TANK , suggesting that MARCH5 regulated the auto-ubiquitination of TRAF6 through TANK . This was further supported by probing TLR7-mediated gene expression ( IL6 and TNFα ) when knocking down TANK and MARCH5 at the same time ( Fig . 6E ) . Taken together , these data indicate that MARCH5 potentiates TLR7 signaling by releasing the inhibitory effects of TANK toward TRAF6 . To determine the importance of mitochondrial localization for MARCH5 function , we generated two mislocalization mutants of MARCH5 . MARCH5 CAAX was constructed by replacing its C-terminal transmembrane domain with a targeting sequence to the plasma membrane , and MARCH5 NLS was constructed by replacing the same transmembrane domain with a nuclear localization sequence ( Fig . 7A ) . As expected , MARCH5 CAAX and MARCH5 NLS were targeted to the plasma membrane and nucleus , respectively ( Fig . 7B middle and lower panel ) . Interestingly , MARCH5 CAAX and MARCH5 NLS failed to interact with TANK , whereas the MARCH5-DM , the C-terminal transmembrane domain truncation , could still bind TANK ( Fig . 7C ) . Corroborating the activity of the mislocalization mutants , neither MARCH5 CAAX nor MARCH5 NLS could catalyze the poly-ubiquitination of TANK ( Fig . 7D , E ) . Furthermore , neither MARCH5 CAAX nor MARCH5 NLS could potentiate the expression of IL6 or TNFα induced by R837 or LPS ( Fig . 7F , G ) . Collectively , these data indicate that mitochondrial localization of MARCH5 is essential for its regulatory function in innate immunity .
A new paradigm has been established in the past decade , revealing how Toll-like receptors ( TLRs ) detect a wide range of pathogens and then initiate immediate host defenses . As a result , cytokines and chemokines are induced to mobilize immune cells for controlling and eliminating pathological infections . Given that the TLR signal transduction cascade is the first line of the host defense against pathogens , they are subjected to multiple layers of positive and negative regulations . Herein , we characterize the mitochondrial protein MARCH5 as an essential and positive modulator of TLR7 signaling . In this study , several lines of evidence substantiate the novel function of MARCH5 in TLR7 signaling . First , exogenous expression of MARCH5 potentiated the induction of NF-κB responsive genes upon R837 stimulation , but not the induction of IRF3/7 responsive genes . Second , knockdown of MARCH5 unequivocally resulted in the reduction of NF-κB-mediated gene expression , and this attenuation was rescued by exogenously expressing a siRNA-resistant rMARCH5 . Third , MARCH5 interacted with TANK , a negative regulator of TLR7 signaling . Interestingly , TANK could partially co-localize to mitochondria in response to TLR7 stimulation . This interaction was increased upon R837 challenge , suggesting that the interaction was transient and dynamic . Fourth , knockdown of TANK impaired the ability of MARCH5 to potentiate TLR7 signaling . Previous in vitro studies suggested that TANK positively regulates TBK1 and IKKε-mediated production of type I interferon [36] . However , analysis of TANK−/− mice indicated that TANK is not essential for the induction of type I interferon downstream of RIG-I/MDA5 or TRIF [19] . Further analysis revealed that TANK is critical for the negative regulation of canonical NF-κB activation via suppression of TRAF6 auto-ubiquitination [19] . The underlying mechanism is still not clear . Ubiquitination is an effective mechanism to regulate TLR signaling pathways . E3 ubiquitin ligases ( Nrdp1 , A20 ) and de-ubiquitinases ( DUBA , CYLD , and A20 ) have been demonstrated as positive or negative modulators of these pathways . Nrdp1 ‘preferentially’ promotes TLR-mediated production of type I interferon [37] . A20 and CYLD ‘preferentially’ terminate TLR-induced activation of NF-κB by de-ubiquitinating their substrates , such as RIP1 , TRAF6 , TAK1 , NEMO and so on [15] , [16] . DUBA interacts with and de-ubiquitinates TRAF3 , thereby attenuating TLR-dependent and TLR-independent antiviral responses [38] . K48-linked poly-ubiquitin chains usually target substrates for proteasome degradation , whereas K63-linked poly-ubiquitin chains usually regulate substrate activity but do not promote effective degradation . Gatot et al . previously reported that TANK is subjected to lipopolysaccharide mediated K63-linked poly-ubiquitination [35] . However , the identity of the relevant E3 ubiquitin ligase and the functional consequence of TANK poly-ubiquitination remained to be revealed . In this study , we showed that ectopic expression of MARCH5 enhanced poly-ubiquitination of TANK after R837 stimulation , whereas knockdown of MARCH5 attenuated this poly-ubiquitination . Wild type MARCH5 catalyzed the K63-linked poly-ubiquitination of TANK on its Lysines 229 , 233 , 280 , 302 and 306 . Neither MARCH5 C2A nor MARCH5 ΔRING could synergize the activation of NF-κB , stimulated by R837 . The siRNA-resistant mutants ( rMARCH5 C2A or rMARCH5 ΔRING ) could not rescue the NF-κB activation in MARCH5 knockdown cells . In addition , TANK per se could interfere with TRAF6 auto-ubiquitination . The in vitro ubiquitination assay revealed that MARCH5 could catalyze the formation of both K48- and K63-linked polyUb chains . This is somewhat unexpected given that MARCH5 mediates only K63-linked polyubiquitination of TANK in transfected cells . One possible explanation for this discrepancy is that additional cofactors may exist in vivo to guide the reaction of MARCH5 in favor of the K63 linkage . Indeed , we have previously shown that TRAF6 only catalyzes the formation of K63-linked polyUb chains in vivo , which positively regulate the NF-κB signaling pathway [39] . Similar to what is observed for MARCH5 , TRAF6 could facilitate the assembly of both K48- and K63-linked polyUb chains in vitro [40] . Recent mechanistic studies of polyUb chain formation revealed that additional proteins were involved in the determination of polyUb linkage [41] , [42] . Strikingly , the NEMO protein is reported to be modified in vivo by linear- , K27- and K63-linked polyUb chains under different physiological conditions [43] . It is thus a great challenge to monitor the dynamic formation of various polyUb chains of different linkages and determine their cognate functional consequences . Currently , we are trying to use the RNAi approach to screen for potential cofactor ( s ) of MARCH5 . Hopefully , this will shed new light on how MARCH5 catalyzes the formation of K63-linked polyUb chains in vivo . Taken together , we propose that MARCH5 is an authentic E3 ubiquitin ligase and catalyzes K63-linked poly-ubiquitination of TANK . MARCH5 modulates TLR7 signaling via releasing the inhibitory action of TANK toward TRAF6 . We speculate that MARCH5 potentiates TLR7 and TLR4 signaling via a similar mechanism . Interestingly , this regulatory function was less potent for TLR4 signaling , probably due to the observation that other TRAFs could partially mediate TLR4 signaling to NF-κB activation [44] . Consistently , it appears that MARCH5 does not influence the activation of IRF3/7 , since TANK displays no regulatory role toward TRAF3 [19] . Interestingly , mislocalization of MARCH5 to either the plasma membrane or the nucleus abolishes its function toward TLR7 signaling . An increasing number of sub-cellular organelles are functionally connected to the anti-microbial defense system . For example , it has been proposed that endosomes and lysosomes contain TLR3/7/8/9 and probably TLR4 [45] . Recognition of PAMPs by TLRs actually takes place inside these intracellular membrane structures , instead of on the plasma membrane . In addition , the endoplasmic reticulum ( ER ) plays an active role during the transport of TLRs to their appropriate locations . Unexpectedly , mitochondria have recently been uncovered as a new platform for sensing intracellular virus infections . Notably , the existence of cross-talk between TLR signaling and mitochondrial proteins remained unknown . Arguably , our current study reveals the first mitochondrial protein to positively regulate TLR signaling . Because mitochondria and the ER are physically connected , we expect that future investigations will uncover more intricate cross-talk between them during TLR signaling . MARCH family proteins contain a RING finger domain and some trans-membrane motifs . Notably , many of the 11 mammalian MARCH proteins have been implicated in modulating immune functions , either directly or indirectly . MARCH8 ( also known as c-MIR , cellular modifier of immune response ) was demonstrated to specifically catalyze B7 . 2 ubiquitination and its subsequent lysosomal degradation . In addition , both MARCH8 and MARCH1 negatively modulate the expression of CD95 ( Fas ) , TfR ( transferrin receptor ) and MHC class II [28] . MARCH4 and MARCH9 influence antigen presentation by MHC class I molecules . Furthermore , ectopic-expression of MARCH9 leads to the down-regulation of surface ICAM1 , a co-stimulatory molecule for T and B cells [28] . It will be intriguing to examine whether other MARCH family proteins play a critical role in immune regulation .
MARCH5 , TANK , GIDE , TRAF6 , TRAF3 , NEMO , MAVS and RIG-I cDNAs were amplified by PCR from thymus cDNA library ( Clontech ) and subsequently cloned into mammalian expression vectors as indicated . His-Ub and the reporter plasmids ( κB-luciferase and PRDIII-I-luciferase reporters ) have been described previously [46] . All point mutations were introduced by using a QuickChange XL site-directed mutagenesis method ( Stratagene ) . All constructs were confirmed by sequencing . Rabbit polyclonal anti-MARCH5 antibody was a gift from Shigehisa Hirose ( Tokyo Institute of Technology , Midori-ku , Yokohama , Japan ) . Anti-TANK antibodies were raised in rabbits against full-length mouse TANK and affinity purified using an antigen column . Other commercially available antibodies and reagents used were as follows: HA , Myc , Tom20 and Ub antibodies were purchased from Santa Cruz Biotechnology , Inc . Tubulin , Flag and β-actin antibodies were obtained from Sigma-Aldrich . R837 and lipopolysaccharide were purchased from Sigma-Aldrich . HEK293T and RAW264 . 7 cells were cultured using DMEM ( Invitrogen ) plus 10% FBS ( Hyclone ) , supplemented with 1% penicillin-streptomycin ( Invitrogen ) . The procedure for generating BMDCs ( bone marrow-derived dendritic cells ) has been described previously [26] . Lipofectamine ( Invitrogen ) was used for transient transfection of HEK293T Cells . RAW264 . 7 cells were transfected with Nucleofector ( Amaxa ) . Small interference RNA was transfected with Lipofectamine 2000 ( Invitrogen ) according to the manufacturer's instructions . For immuno-precipitation analysis , cells were lysed in TBS buffer ( 50 mM Tris-Cl pH 7 . 4 , 150 mM NaCl ) supplemented with 1% Triton-X 100 , 1 mM PMSF and a protease inhibitor cocktail ( Roche ) . After pre-clearing for 1 hour , lysates were incubated with the appropriate antibody for four hours to overnight at 4°C . Two hours after adding protein A/G agarose , the immuno-precipitates were extensively washed with lysis buffer and eluted with SDS loading buffer by boiling for 3 min . For immuno-blot analysis , the samples were resolved by SDS-PAGE and transferred to a PVDF membrane ( Millipore ) . Immunoblotting was probed with indicated antibodies . The proteins were visualized by using a NBT/BCIP Western blotting system ( Promega ) or a SuperSignal West Pico chemiluminescence ECL kit ( Pierce ) . For Ni-NTA pulldown analysis , cells were washed with PBS and then lysed in His-Lysis Buffer ( 50 mM Tris-Cl . pH 7 . 4 , 6M Urea ) . 20 µL Ni-NTA agarose beads ( Qiagen ) were then added into the post-centrifuged lysates and rotated for four hours at 4°C . After extensively washing with His-Lysis Buffer , the precipitates were boiled with SDS loading buffer and then subjected to SDS-PAGE followed by immuno-blot analysis . Cells were seeded in 12-well plates and transfected with reporter gene plasmids combined with siRNAs and other constructs as indicated . The total amount of DNA was kept constant by supplementing with empty vectors . pTK-Renilla was cotransfected to normalize transfection efficiency . Luciferase activity was analyzed with the Dual Luciferase Reporter Assay System ( Promega ) . Total RNA was extracted using TRIzol reagent ( Invitrogen ) according to the manufacturer's instruction . Reverse transcription of purified RNA was performed using oligonucleotide dT primer . The quantifications of gene transcripts were performed by real-time PCR using Power SYBR GREEN PCR MASTER MIX ( ABI ) . All values were normalized to the level of β-actin mRNA . The primers used are as follows: Chemically synthesized 21-nucleotide siRNA duplexes were obtained from Gene-Pharma and transfected using Lipofectamine 2000 ( Invitrogen ) according to the manufacturer's instructions . RNA oligonucleotides used in this study are as follows: Cells were fixed with 4% paraformaldehyde , permeabilized with 0 . 1% Triton X-100 , blocked with 1% bovine serum albumin and stained with the indicated primary antibodies followed by FITC-conjugated secondary antibody ( Jackson Immuno-Research Laboratories ) . Nuclei were counterstained with DAPI ( Sigma-Aldrich ) . For mitochondria staining , living cells were incubated with 300 nM Mito Tracker Red ( Invitrogen ) for 30 min at 37°C . Slides were mounted by Aqua-Poly/Mount ( Polysciences ) . Imaging of the cells was carried out using Leica laser scanning confocal microscopy . Preparations of cell suspensions for cell cycle analysis were performed as previously described [47] . Briefly , cells were trypsinized and fixed in ice-cold 70% ethanol at 4°C . After being washed with PBS twice , DNA was stained with 20 µg/mL propidium iodide ( Sigma-Aldrich ) in the presence of 200 µg/mL RNase A ( Fermentas ) . Data were acquired on a FACSCalibur ( BD Biosciences ) and analyzed using CellQuest ( BD Biosciences ) and FlowJo ( TreeStar Inc . ) . For the synthesis of polyUb chains , purified MBP-MARCH5-NT ( 0 . 1 µM ) was mixed with E1 ( 50 nM ) , E2 ( 0 . 3 µM ) , ubiquitin or ubiquitin mutants ( 0 . 1 mM ) ( Boston Biochem ) in a reaction buffer containing 50 mM Tris–HCl , pH 7 . 5 , 5 mM MgCl2 , 2 mM ATP , 2 mM DTT . The reaction was carried out at 30°C for 90 min and then resolved by SDS-PAGE . Ubiquitinated products were detected by immunoblotting with a Ub-specific antibody ( P4D1 ) . Concentrations of cytokines in culture supernatants were measured by ELISA kits ( R&D Systems ) according to the manufacturer's instructions . HEK293T cells were washed with cold PBS and lysed by douncing in homogenization buffer ( buffer H: 210 mM sucrose , 70 mM mannitol , 1 mM EDTA , 1 mM EGTA , 1 . 5 mM MgCl2 , 10 mM Hepes ( pH 7 . 2 ) , protease inhibitor cocktail ) . The homogenate was centrifuged at 500×g for 10 min , and the pellet ( P1 ) was saved as crude nuclei . The supernatant ( S1 ) was centrifuged at 5 , 000×g for 10 min to precipitate mitochondria ( P5 ) . The whole cell and fractions of P5 and S5 were lysed in lysis buffer ( 50 mM Tris-HCl . pH 7 . 4 , 150 mM NaCl , 1 mM EDTA , and 1% Triton-X 100 , protease inhibitor cocktail ) followed by immunoblot analysis . Student's t test was used for the statistical analysis of two independent treatments . For all tests , a P value of <0 . 05 was considered statistically significant . The GenBank ( http://www . ncbi . nlm . nih . gov/Genbank ) accession numbers for the genes and gene products discussed in this paper are: MARCH5 ( NM_027314 , NP_081590 ) , TANK ( NM_001164071 . 1 , NP_001157543 . 1 ) , GIDE ( NM_026689 . 3 , NP_080965 . 2 ) , TRAF6 ( NM_004620 . 2 , NP_004611 . 1 ) , TRAF3 ( NM_145725 . 2 , NP_663777 . 1 ) , NEMO ( NM_001099857 . 1 , NP_001093327 . 1 ) , MAVS ( NM_020746 . 4 , NP_065797 . 2 ) , RIG-I ( NM_014314 . 3 , NP_055129 . 2 ) , TRIF ( NM_182919 . 2 , NP_891549 . 1 ) , MyD88 ( NM_002468 . 4 , NP_002459 . 2 ) , IRAK1 ( NM_001025242 . 1 , NP_001020413 . 1 ) .
|
In 2005 , MAVS was characterized as the critical adaptor protein for the signal transduction of RIG-I-like receptors ( RLRs ) . This provided the first link between mitochondria and the intracellular antiviral defense system . From then on , exploring the potential functions of novel mitochondrial proteins in microbe-host interactions became a rapidly expanding frontier . Notably , it remains unknown whether mitochondrial proteins can directly regulate TLR signaling . Here , we demonstrate that the mitochondrial protein MARCH5 positively modulates TLR7 signaling . Our study reveals that MARCH5 is a novel E3 ubiquitin ligase and catalyzes the K63-linked poly-ubiquitination of TANK . This modification releases the inhibitory effects of TANK on TRAF6 . Arguably , this represents the first study linking mitochondria to TLR signaling , shedding new light on the role of mitochondria in the proinflammatory response .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"molecular",
"cell",
"biology",
"immunity",
"immunology",
"biology",
"immune",
"response"
] |
2011
|
Mitochondrial Ubiquitin Ligase MARCH5 Promotes TLR7 Signaling by Attenuating TANK Action
|
The establishment of the germline is a critical , yet surprisingly evolutionarily labile , event in the development of sexually reproducing animals . In the fly Drosophila , germ cells acquire their fate early during development through the inheritance of the germ plasm , a specialized maternal cytoplasm localized at the posterior pole of the oocyte . The gene oskar ( osk ) is both necessary and sufficient for assembling this substance . Both maternal germ plasm and oskar are evolutionary novelties within the insects , as the germline is specified by zygotic induction in basally branching insects , and osk has until now only been detected in dipterans . In order to understand the origin of these evolutionary novelties , we used comparative genomics , parental RNAi , and gene expression analyses in multiple insect species . We have found that the origin of osk and its role in specifying the germline coincided with the innovation of maternal germ plasm and pole cells at the base of the holometabolous insects and that losses of osk are correlated with changes in germline determination strategies within the Holometabola . Our results indicate that the invention of the novel gene osk was a key innovation that allowed the transition from the ancestral late zygotic mode of germline induction to a maternally controlled establishment of the germline found in many holometabolous insect species . We propose that the ancestral role of osk was to connect an upstream network ancestrally involved in mRNA localization and translational control to a downstream regulatory network ancestrally involved in executing the germ cell program .
Germ cells are essential for the transfer of heritable information and , therefore , the determination of their fate is a critical event in the development and evolution of sexually reproducing organisms . Two general strategies for generating the germline have evolved in animals: cytoplasmic inheritance or zygotic induction . Inheritance requires that determinants of the germ cell fate ( mRNAs and proteins that form the pole plasm ) are maternally generated and provisioned to the oocyte . In contrast , induction involves the acquisition de novo of the germ cell fate in a subset of cells later during embryonic development [1] , [2] . Some of the first experiments that proved the existence of a maternally generated substance capable of inducing the germline fate were conducted in insects . It had been observed that in many insect species , a distinct region of cytoplasm ( called pole plasm , or oosome ) is localized to the posterior pole of the oocyte during oogenesis . This pole plasm remains at the posterior during early embryogenesis , until cleavage nuclei reach the embryo cortex . Those nuclei that reach the posterior pole of the embryo interact with the pole plasm , bud from the posterior pole , and become cellularized precociously in comparison to the other blastodermal nuclei [3] . These cells are termed pole cells , and will give rise to the germline [4] , [5] . Classical embryonic manipulations showed that the pole plasm is both necessary [6] , and sufficient [7] to produce the primordial germ cells . Genetic analyses have identified numerous molecular factors that are required for the proper production of the pole plasm and pole cells in Drosophila . Only one of these , oskar ( osk ) , is both necessary and sufficient to induce the production of polar granules and pole cells [8] . Due to the sufficiency of Osk to induce germ plasm , it must be tightly regulated to prevent ectopic induction of germline fate . To this end , genes upstream of osk are generally required to regulate translation of osk mRNA and to mediate its transport between the time it is transcribed in the nurse cells and the time it is properly posteriorly localized in the oocyte [9] . Genes downstream of osk are generally required to assemble the polar granules or to mediate proper behavior of the pole cells [9] , and have highly conserved functions in the germline throughout the Metazoa [10]–[12] . Current data suggest that the mode of germline determination found in Drosophila is not the ancestral mode among the insects . So far neither unequivocal maternal germ plasm nor pole cells have been detected in representatives of basally branching hemimetabolous insect orders . Rather , species from these orders instead appear to rely on zygotic induction mechanisms to specify their germline [13]–[17] ( Figure 1 ) . Consistent with absence of cytoplasmic inheritance of germline determinants and the production of pole cells , the processes for which osk is required , orthologs of osk have not been detected in any of the sequenced genomes of the hemimetabolous insects Acyrthosiphon pisum [18] , Rhodnius prolixis ( http://genome . wustl . edu/genomes/view/rhodnius_prolixus/ ) , and Pediculus humanus http://phumanus . vectorbase . org/SequenceData/Genome/ ( Figure 1 , Table S1 ) . Among the Holometabola , osk orthologs are also apparently absent from the sequenced genomes of the silk moth Bombyx mori ( Lepidoptera ) [19] , the beetle Tribolium castaneum ( Coleoptera ) [20] , and the honeybee Apis mellifera ( Hymenoptera ) [21] ( Figure 1 , Table S1 ) . Consistent with this absence osk , Bombyx , Tribolium , and Apis all also lack maternal germ plasm , do not produce pole cells , and appear to rather use zygotic inductive strategies to generate the germline [22]–[25] ( Figure 1 ) . These observations led to the idea that osk may have been a novelty that originated within the dipteran lineage [26] , [27] . However , Drosophila-like modes of germline determination through posteriorly localized maternal germ plasm and pole cells are also found throughout the Holometabola , including most major lineages of the Hymenoptera ( e . g . , Nasonia vitripennis [28] sawflies [29] and multiple ant species [30] , [31] ) , the Coleoptera ( e . g . , Acanthoscelides obtectus [32] , Dermestes frischi [33] ) , Megaloptera ( Sialis misuhashii [34] ) and Lepidoptera ( Pectinophora gossypiella [35] ) ( Figure 1 ) . Despite the similarity of the strategies for germline determination in the above species to that employed in Drosophila , osk orthologs have only been identified in the genomes of the dipterans Anopheles gambiae , Aedes aegypti , and Culex pipiens [36] , [37] ( Figure 1 ) . These observations raised the question of evolutionary origin of osk in the insects and whether or not this gene is associated with the evolution of the inheritance mode of germline specification . To answer these fundamental questions , we examined the molecular basis of maternal germ plasm production in the wasp Nasonia vitripennis . We chose Nasonia because its genome was recently sequenced [38] , it is amenable to functional manipulation by pRNAi [39] , and its key phylogenetic position within the most basally branching holometabolous order , the Hymenoptera [40] , [41] . We show that the regulatory network underlying the production of maternal germ plasm and pole cells is largely conserved between Nasonia and Drosophila , and argue that these features had a common phylogenetic origin at the base of the Holometabola . In addition , we provide evidence that the possession of an oskar ortholog is a general feature of insects that produce pole cells , and that oskar has likely been lost independently multiple times within the Holometabola in correlation with shifts in strategies for establishing the germline .
Attempts to detect a Nasonia ortholog by BLAST [42] searches using the Drosophila Osk sequence as the query failed to return significant hits . However , using Oskar sequences identified in the mosquitoes Culex and Aedes , we identified a Nasonia genomic region that showed significant similarity to the mosquito sequences . Using the predicted peptide sequence in this region , reciprocal BLAST against the mosquito and Drosophila genome databases returned results with significant E-values that corresponded to osk genes in each of these species ( Table S1 ) . We thus hypothesized that the region in the wasp genome detected by mosquito Osk BLASTs corresponded to Nasonia osk , and cloned a 1500 base pair fragment representing the full length complementary DNA of Nasonia osk using RACE PCR . This sequence contains an open reading frame that is predicted to generate a protein of 375 amino acids . The overall Nv-Osk sequence is similar to that of Drosophila Osk ( 16% identity , 33% similarity , 44% gaps ) , and many of the residues critical for fly Osk function are conserved in the Nasonia sequence ( Figure 2 ) . However , we could identify two regions that appear to be unique to the fly sequence . One is the region that is specific to the Drosophila long-Osk isoform [43] ( Figure 2 , red text ) . No similarity to this region appears to be encoded in the Nv-osk mRNA , nor is it present in mosquito Osk sequences . The other region that is absent in Nv-Osk includes amino acids 290 to 396 in Dm-Osk ( Figure 2 , blue text ) , which corresponds to the domain interacting with LASP to regulate Osk anchoring to the actin cytoskeleton [44] . Interestingly , this region is also absent from the mosquito Osk sequences , which appear to be more similar to Nv-Osk in sequence and general structure ( Culex/Nasonia: 24% identity , 42% similarity , 22% gaps ) . A search in the Conserved Domain Database indicates that the central portion of the Nv-Osk protein shares similarity with a GDSL/SGNH-hydrolase or lipase-like domain ( Figure 2 , orange boxes ) , consistent with similar observations made for C . pipiens and A . aegypti Osk orthologs [36] . This domain is weakly detected in Drosophila Osk and it is not clear whether it is necessary for Osk function in pole plasm assembly . In addition , the N-terminal region of Nv-Osk shows strong similarity to a domain also present at the N-termini of highly conserved tudor-domain containing proteins . This domain has been independently identified in silico as either the Lotus domain [45] , or Tejas domain [46] . This domain is present at the N-terminus of orthologs of tudor-domain-containing-7 and -5 ( tdrd7 , tdrd5 ) , and related tudor domain containing genes [47] , and is detected only weakly in fly Osk . tdrd7 and tdrd5 orthologs are found throughout the Metazoa , including all sequenced insect genomes ( JAL , personal observation ) , and are characterized by the presence of Tudor domains toward the C-terminus of the protein , which are absent in Osk proteins . The N-terminal 100 amino acids of Nv-Osk show strong homology to Tdrd7 orthologs throughout the Metazoa , ranging from 39% identical ( BLAST E-value 8e-09 ) to the Apis ortholog , 31% identical ( BLAST E-value 1e-05 ) to the Hydra ortholog , and 29% identical ( BLAST E-value 7e-05 ) for the Danio ( zebrafish ) ortholog . In comparison , the Apis and Danio Tdrd7 C-termini are 49% identical ( BLAST E-value 2e-13 ) , and Apis and Hydra proteins are 30% identical ( BLAST E-value 3e-09 ) in the N-terminal region . In zebrafish , tdrd7 has a role in controlling germ granule morphology and number during embryogenesis [48] . Furthermore , the Drosophila tdrd5 ortholog , tejas , has a critical role in germline development , and the N-terminal region of this protein ( including the Tejas domain , which is similar to the N-terminus of Nv-Osk ) has been shown to physically interact with Vas [46] . Finally , a bioinformatic analysis of proteins containing domains similar to those found in Osk and Tdrd7/5 N-termini ( termed by the authors OST-HTH ) indicated that these domains may bind double-stranded RNA [49] . These results indicate that Oskar is at least partially related to genes that had ancestral germline and/or RNA binding functions . Nasonia oogenesis occurs in ovarioles of the polytrophic-meroistic type , where each oocyte is associated with its own population of nurse cells , and has been described in detail previously [50] . Nv-osk mRNA is detected quite early in oogenesis , just after the time that the nurse cells become distinguishable from the oocyte ( Figure 3A , 3A′ ) . As the egg chambers mature ( Figure 3B ) , Nv-osk is expressed at very high levels in only the posterior nurse cells nearest to the oocyte . Within these cells , Nv-osk mRNA is incorporated into particles ( Figure 3B ) , a pattern similar to that of Nv-otd1 [51] . From the very early stages of oogenesis , Nv-osk is transported from the nurse cells to the oocyte , where it is localized to the posterior pole in a pattern similar to that of Nv-nos ( Figure 3A′ , 3B , 3C ) . During late oogenesis , Nv-osk mRNA levels go from high to barely detectable in the nurse cells of adjacent egg chambers ( Figure 3C ) . This likely indicates the onset of nurse cell dumping , as from this point on the nurse cells will become progressively smaller and eventually disappear . This pattern of rapid transfer of mRNA is similar to what is seen for Nv-otd1 during late oogenesis , except that Nv-otd1 mRNA accumulates at the anterior pole of the oocyte at this stage [51] . In the early embryo , Nv-osk mRNA remains localized to the posterior pole , and most of the mRNA is associated with the oosome , a large , discreet structure associated with the posterior pole . The oosome migrates within the embryo during the early cleavages ( Figure 3D ) , before returning to the posterior pole just before the formation of pole cells ( Figure 3E , see [51] for details ) . At this stage , a population of Nv-osk mRNA not contained within the oosome is observed in a gradient at the posterior pole , a pattern which is typical for oosome associated mRNAs ( e . g . , otd1and nanos in Nasonia [52] ) . Nv-osk mRNA still associated with the oosome is then incorporated into the pole cells ( Figure 3F ) , while the cytoplasmic population remains in the embryo proper ( not shown , but see [51] for expression of Nv-nos mRNA , which shows identical behavior at these stages ) . Both populations of mRNA are finally degraded as the cellular blastoderm begins to form ( Figure 3G ) . We used parental RNA interference ( pRNAi ) to analyze the function of Nv-osk during Nasonia development . We obtained specific phenotypes that vary in terms of intensity allowing us to infer a number of potential functions for Nv-osk during oogenesis and early embryogenesis . In ovarioles showing the strongest Nv-osk pRNAi effect , only a few egg chambers are produced ( Figure 4B , compare to 4A ) indicating that Nv-osk has an early role in promoting oogenesis . This may be related to a similar phenotype produced by mRNA null mutations in fly osk [53] . The Nasonia ovariole normally consists of a linear array of egg chambers , with the oocytes always lying directly posterior to their sister nurse cells and directly anterior to the next older egg chamber ( Figure 4A ) . In the milder phenotypes of Nv-osk pRNAi , this linear arrangement is disrupted , and egg chambers arranged perpendicularly to the long axis of the ovariole ( arrows in Figure 4C and 4D ) , or with reversed polarity ( arrowhead Figure 4C ) are observed . Egg chamber polarity defects are also observed after pRNAi against Nv-vas ( not shown ) and Nv-tud ( see below ) , indicating that there is a novel role for germ plasm components in establishing polarity of egg chambers within the ovarioles of Nasonia . Due to the variability in the final morphology of ovarioles after pRNAi for Nv-vas , -osk , and -tud , it is not clear whether these phenotypes are all the result of the disruption of a single developmental process . Within the oocytes , Nv-nos and otd1 mRNAs are sometimes localized more loosely than normal ( asterisk and arrowhead Figure 4C ) or mislocalized in relation to the AP axis of the oocyte ( asterisk Figure 4D ) after Nv-osk pRNAi . These phenotypes may represent a disruption of the internal polarity of the oocytes and/or proper anchoring of localized mRNAs . A more detailed understanding of oocyte cytoskeletal polarity and mRNA anchoring mechanisms in Nasonia will be required to resolve this uncertainty . In any case , these results indicate that Nv-osk is required for germline development , for establishing the polarity of the egg chambers , and for the proper localization of the pole plasm to the posterior pole . In Drosophila , the recruitment of Vas protein to the posterior pole of the oocyte by Osk is a critical step in polar granule assembly . To test whether Nv-Osk functions in a similar way , we examined the distribution of Nv-Vas using a Nasonia specific Vasa antiserum in wild type and Nv-osk pRNAi ovaries . During early oogenesis , Nv-Vas protein is detected primarily on the surface of the nuclei of the most anterior nurse cells ( Figure 4E′ ) . This is consistent with the strong transcription of Nv-vas detected in these cells ( Figure S1A ) . Localized Nv-Vas protein is not seen in early oocytes ( Figure 4E′ ) , even though Nv-osk is already localized at high levels at the posterior ( Figure 4E ) . Localized Nv-Vas becomes visible in the oocyte relatively late in oogenesis , when the oocyte is of the same size as the nurse cell cluster ( Figure 4F , 4F′ ) . This accumulation of Nv-Vas at the posterior pole is abolished after Nv-osk pRNAi ( Figure 4G ) , while Nv-Vas production in anterior nurse cells appears unaffected ( Figure 4G′ ) . Thus , the role of Osk in recruiting germ plasm components to the posterior pole is conserved between Drosophila and Nasonia . Posteriorly localized mRNAs ( e . g . , Nv-nos , Nv-otd1 and Nv-osk ) are incorporated into the oosome in early Nasonia embryos ( Figure 5A ) . After Nv-osk pRNAi , these mRNAs remain in a homogenous cap at the posterior pole of the embryo , and the oosome is not formed ( 100% penetrance , N = 60 ) ( Figure 5B , 5C ) . In addition , the anterior localization of Nv-otd1 mRNA is disrupted . Rather than being tightly localized at the anterior pole , Nv-otd1 mRNA is often seen in particles distributed throughout the anterior half of the embryo ( Figure 5B ) . This part of the phenotype may be related to the polarity defects observed in Nv-osk pRNAi oocytes . pRNAi against Nv-osk also results in the completely penetrant ( N = 57 ) loss of pole cells ( Compare wild type in Figure 5D to 5E ) . In the absence of the protective environment of the pole cells , all Nv-nos mRNA is lost from the embryo by the late blastoderm stage ( Figure 5F ) . A similar phenomenon is seen after Nv-vas pRNAi [51] . Nv-osk pRNAi also causes embryonic patterning phenotypes that result in larval lethality ( 42% , N = 75 ) . Only a portion ( 13% ) showed phenotypes similar to Nv-nos pRNAi [51] , while the remainder of affected cuticles showed defects in head patterning , or more severe patterning disruptions of unclear origin . This range of phenotype was also seen for Nv-vasa [51] , and these observations indicate that the roles of Nasonia germ plasm assembly factors in embryonic patterning are much more complicated than they are in the fly , where nos mRNA translation is the main embryonic patterning output of germ plasm assembly [54] . In Drosophila , Oskar acts through two main downstream proteins to produce polar granules: Vas and Tud [9] . As shown above , Nv-Osk functions upstream of Nv-Vas recruitment to the posterior during oogenesis ( Figure 4G′ ) . However , the functional relationship between Nv-Osk and Nv-Vas in the ovary may not be strictly hierarchical , as Nv-Vas knockdown ( Figure 6A′ ) leads to defects in the proper anchoring and tight localization of Nv-osk mRNA to the posterior pole of the oocyte ( Figure 6A ) . In the embryo , Nv-vasa pRNAi results in the completely penetrant loss of the oosome ( Figure 6E ) and pole cells ( Figure 6F ) , similar to the effects of Nv-osk pRNAi . In contrast to Nv-osk and Nv-vas pRNAi , knockdown of Nv-tud , which is expressed weakly and ubiquitously in the nurse cells and oocyte ( Figure S1B ) , has only a minor effect on posterior accumulation of Nv-Vas protein in the oocyte , even when strong polarity defects within the ovariole are observed ( Figure 6B , 6B′ ) . In the embryo , the oosome is still formed , but is significantly reduced in size ( Compare Figure 6G to 6C ) . In line with these apparently weaker effects , Nv-tud pRNAi leads to a reduction in the number of pole cells , and those that do form are smaller , less spherical , and less segregated from the somatic nuclei at the posterior pole which may indicate that they are not completely differentiated as primordial germ cells ( Compare Figure 6H to 6D ) . These results indicate that , similar to fly tud [8] , [55] , Nv-tud function is downstream of Nv-vas and Nv-osk in the production of the germ plasm . However , due to the incompleteness and variability of pRNAi efficiency , we cannot exclude the possibility that the weaker defects are the result of general weaker knockdown of Nv-tud with pRNAi . In Drosophila , the localization and regulation of osk translation is tightly regulated in order to prevent ectopic pole plasm and disruptions in segmental patterning . A critical factor in ensuring proper control of osk translation is the RNA binding protein Bruno , which binds the UTRs of osk mRNA and represses its translation . This repression is relieved under normal circumstances only upon localization of osk mRNA to the posterior pole of the oocyte [56] . We analyzed the function of Nasonia bruno to test whether a similar mechanism of translational repression operates in Nasonia to prevent the ectopic assembly of the oosome . In wild-type egg chambers , Nv-osk and otd1 mRNAs are co-expressed in the posterior nurse cells and localized at the posterior pole of the oocyte , while Nv-otd1 is additionally localized to the anterior pole ( Figure 7A , 7A′ ) . The distribution of these mRNAs is dramatically altered after Nv-bruno RNAi: both Nv-osk and Nv-otd1 ( and Nv-nos , data not shown ) mRNAs are concentrated in large , dense , spheroid particles in the posterior-most nurse cells ( Figure 7B , 7B′ ) . These large particles seem to originate at the nuclear envelope , and smaller particles are observed on the surface of the nurse cell nuclear membranes in some egg-chambers ( Figure 7C , 7C′ ) . The morphology ( density , large size , spheroidal shape ) and molecular composition of the ectopic particles seen after Nv-bruno RNAi are similar to the corresponding features of the oosome , indicating that this structure is being ectopically produced in the nurse cells . If the role of Nv-bruno is similar to that of its Drosophila ortholog , the production of these oosome-like structures in the nurse cells could be due to the ectopic translation of Nv-osk in the nurse cells in the absence of Nv-bruno . In support of this conclusion , the large particles are only produced in the most posterior nurse cells nearest to the oocyte , to which Nv-osk is restricted ( Figure 3 ) , while Nv-bruno is expressed in nurse cells located more anteriorly ( Figure S1C ) . However , we cannot exclude that the restriction of large oosome-like particles to the posterior nurse cells is a result of higher levels of Nv-Bruno protein in these cells . In addition , in late Nv-bruno pRNAi egg chambers , Nv-Vas protein is associated with the dense accumulation of Nv-osk mRNA ( Figure 7D ) , further indicating that oosome formation is being completed ectopically within the nurse cells . Conclusive evidence for a direct role of Nv-Bruno in repressing Nv-osk translation will come only with the availability of an antibody against Nv-Osk protein . Another Drosophila RNA binding protein , Hrp48 , is critical for both silencing of unlocalized osk mRNA translation , and for the proper initiation of its translation once the mRNA is localized to the posterior [57] , [58] . Nv-hrp48 is expressed strongly throughout the nurse cells in the wasp ovary ( Figure S1D ) , and when its function is knocked down , ectopic oosome-like structures are not seen in the nurse cells ( Figure 7E , 7F ) , in contrast to what is seen after Nv-bruno pRNAi . In most egg chambers , both Nv-osk and Nv-otd1 mRNAs are expressed normally in the nurse cells , and are transported to the oocyte ( Figure 7E ) . Once in the oocyte , however , these mRNAs do not become localized normally . The extent of mislocalization varies from oocytes that show a looser localization of posterior mRNAs ( Figure 7E ) to those where Nv-osk and Nv-otd1 mRNAs fail to localize to a distinct cortical location , and are diffusely expressed throughout the smaller than usual oocytes ( Figure 7F , arrow ) . In more weakly affected egg chambers , which have established normal polarity , the pattern of Nv-Vas accumulation appears to be only weakly affected , with the protein appearing at slightly lower levels , and loosely organized , likely reflecting a mild disruption in the proper assembly of the oosome during late oogenesis ( Figure 7G , 7G′ ) . Thus , Nv-hrp48 appears to have a conserved role in the assembly of the germ plasm in Nasonia , and by extension may have a conserved function in regulating the translation of Nv-osk . Our results indicate that the primary role of this factor is to promote oosome assembly ( and thus , by analogy to Drosophila , Nv-osk function ) . However , we cannot completely exclude a second role , such as that seen in Drosophila , for Nv-hrp48 in Nasonia in repressing the translation of unlocalized Nv-osk in the oocyte [57] , [58] . Our results show that a regulatory network of protein interaction centered on Nv-Osk is required for the maternal production of germ plasm , and that this network is highly similar to that found in Drosophila . This suggests that , given the basally branching phylogenetic position of the Hymenoptera among the Holometabola , this regulatory network arose in a common ancestor of all Holometabola , and that transitions to the zygotic induction mode of germ cell specification are associated with secondary disruptions of this network . To test this hypothesis , we sought to determine if osk , as the central component of this network , is conserved in other species that produce maternal germ plasm and pole cells . Multiple ant species have been shown to specify their pole cells through the assembly of a posterior pole plasm that is incorporated into pole cells during early embryogenesis [30] , [31] . Consistent with our hypothesis , we successfully cloned an osk ortholog in the ant Messor pergandei , whose protein sequence shows 46 . 4% similarity to that of Nv-Osk . Moreover , Messor osk ( Mp-osk ) mRNA is localized to the posterior pole of the oocyte during oogenesis ( Figure 8A ) , and embryogenesis ( Figure 8B ) . This pattern of Mp-osk mRNA accumulation is similar to that of insects that specify germ cell through cytoplasmic inheritance ( e . g . , Nasonia and Drosophila ) , and suggests that its function in germ cell specification is conserved in ants . In addition , the localization of Mp-osk corresponds well to the previously observed localization of Vasa protein and nanos mRNA in the oocyte and embryo at equivalent stages in Messor and other closely related ant species [30] , [31] . Messor is a much closer relative of Apis than is Nasonia [59] , and the discovery of osk in this ant species strongly indicates that the absence of osk in the bee genome is a derived state . We also analyzed the molecular basis of maternal germ plasm formation in the beetle Acanthoscelides obtectus , which , like Nasonia , but unlike Tribolium , produces an oosome and pole cells [32] . Like Tribolium and many other beetle species , Acanthoscelides possesses teletrophic ovarioles . In this type of oogenesis , a common pool of nurse cells is located at the anterior of the ovariole , which is connected with progressively maturing oocytes toward the posterior by actin and microtubule-rich structures called trophic cords [60] . In early oogenesis , Vas protein is highly enriched around the surface of the oocyte nucleues ( Figure 8C ) . The presence of Vas protein is also detected in the nurse cells and trophic cords . In more mature oocytes , Vas protein is strongly enriched at the posterior pole , where the oosome will be formed ( Figure 8D ) . This indicates that , despite employing a mode of oogenesis quite divergent from that seen in Nasonia and Drosophila , this beetle possesses similar capabilities for directing the localization and assembly of the germ plasm components to the posterior pole . This is in contrast to Tribolium , where Vas protein is never found in a localized pattern in later oocytes ( Figure 8F ) despite its presence in the cytoplasm of early oocytes and in the trophic cords ( Figure 8E ) , correlating well with the absence of pole cells and maternal germ plasm in this species . Based on the similarity of the pattern of Vasa protein accumulation in Acanthoscelides to the osk dependent Vas localization patterns in Nasonia and Drosophila , we predict that an osk ortholog is present in the genome of Acanthoscelides , and that it functions in recruiting Vas protein to the posterior pole of the oocyte and in assembling the oosome similar to its orthologs in Nasonia and Drosophila . Attempts to clone osk from the beetle by degenerate PCR have so far failed , and transcriptome or genome sequencing may be required to resolve this question .
Taken together , our results reveal a new picture for the origin and evolution of oskar , maternally provisioned germ plasm , and pole cells . We propose that the origins of these features represent evolutionary novelties of the Holometabola in relation to the rest of the insects , and that the appearance of the latter two features is strongly correlated with the presence of osk ( Figure 9 ) . Our conclusions are based on: ( 1 ) the presence of osk orthologs in the genomes of Nasonia and Messor , two distantly related hymenopteran species that also both have maternal germ plasm and pole cells; ( 2 ) the molecular and developmental similarity of the germ plasm of Acanthoscelides to that of Drosophila and Nasonia , which is consistent with the presence of an osk ortholog in this beetle; ( 3 ) the conserved interactions of Nv-Osk with upstream regulators ( such as Nv-Bruno and Nv-Hrp48 ) and downstream partners ( such as Nv-Vas and Nv-Tud ) , which indicate that a protein interaction network centered on Osk for generating maternal germ plasm and pole cells was present at the latest in the most recent common ancestor of the Hymenoptera and Diptera ( which , based on current phylogenies would also be the common ancestor of all Holometabola ) ( Figure 9 ) ; and finally ( 4 ) the absence of maternal germ plasm , pole cells and osk in hemimetabolous insects , suggesting that the absence of these features is ancestral for the insects ( Figure 9 ) , and that these features likely arose after the divergence of the Holometabola from its sister group the Paraneoptera ( true bugs , lice , and thrips ) . The mapping of our findings on the insect phylogeny also indicates that Apis , Tribolium , and Bombyx may have lost these characters through independent evolutionary events ( Figure 9 ) . In addition , the correlation of the loss of maternal germ plasm and pole cells with the absence of oskar in these species ( Figure 9 ) , indicate that osk is a key factor in the evolution of germline determination mechanisms in the Holometabola . Since production of the germline is a critical event in development and evolution , it is surprising that dramatic changes in how this cell fate is established have occurred several times in insect evolution . Such transitions could have been facilitated if redundant mechanisms for generating the germline existed in the ancestors of lineages that eventually lost the ability to maternally specify the germline . In Drosophila there appears to be no remaining inductive capability: if pole cells are not produced , or are destroyed before reaching the gonad , the resulting fly is sterile . However , this is not the case in all insects . Destruction or removal of the oosome from the embryo of the wasp Pimpla turionellae resulted in the complete absence of pole cells , consistent with the role of the oosome in generating these cells . In spite of this , when embryos subject to these manipulations were examined later , a majority appeared to have germ cells populating the late embryonic gonads [61] . As Pimpla is a close relative of ants and bees , it is possible that both maternally provisioned germ plasm and the ability to zygotically induce germline fate coexisted in an ancestor of Apis , and the loss of the former capability thus may not have had dire consequences for the fecundity of species within the lineage leading to Apis . Once the presence of pole cells and maternal germ plasm was no longer selected for , it may have been relatively easy to lose osk , as long as another strategy for either localizing posterior nanos , or another mechanism for patterning the posterior is present . The question of why an insect would lose the capacity to produce pole cells is also difficult to address directly . The likelihood that maternal provisioning of germline determinants evolved independently multiple times among animals [1] , [2] implies that this strategy for germline determination has , at least under certain circumstances , selective benefits . Reciprocally , the multiple independent losses of this strategy indicate that , in other circumstances , zygotic induction may be favored . Broader sampling of germline specification strategies among the animals could shed light on the possible ecological or embryological traits correlated with the retention of or transition away from maternal synthesis of germline determinants and early segregation of the germ cell fate . Our finding that Oskar was a critical innovation for the transition to the maternal inheritance mode of germline determination in insects leads to the question of how such a novel protein could have been invented . The strong similarity of the N-terminus of Nv-Osk to the N-terminus of Tdrd-7 orthologs found throughout the Metazoa , indicates that the origin of osk involved the duplication and divergence of this locus in an ancestor of the Holometabola . However , unlike tdrd-7 genes , osk orthologs lack Tudor domains toward the C-terminus , and rather have a domain with structural similarity to SGNH/GDSL class hydrolases . Since such a domain is not found in Tdrd-7 orthologs , it may be that osk arose by a fusion of a tdrd7 paralog , and a gene possessing a hydrolase domain . While proteins of the SGNH/GDSL hydrolase family are found in all insect species , Osk orthologs show no significant homology to these sequences in BLAST analyses ( E-value cutoff = 10 ) . Rather , the highest scoring non-Oskar BLAST hits for the C-terminal portion ( i . e . , excluding the first 100 amino acids ) of Osk proteins are often SGNH/GDSL hydrolases of Bacteria ( e . g . , Mp-Osk finds ZP_05979902 . 1 from Subdoligranulum variabile at an E-value of 0 . 17 , and Cp-Osk finds YP_001491067 . 1 from Arcobacter butzleri at an E-value of 0 . 006 ) . These observations raise the possibility that osk could have arisen by the combination of horizontal gene transfer from bacteria and gene fusion events . The fact that horizontal gene transfer from endosymbiotic bacteria occurs in insects is now well established [38] , [62] , and a source for a potential horizontal transfer could be the endosymbionts that are tightly associated with the early germ cells and gonads of many insect species ( e . g . , [63] , [64] ) . While the most parsimonious explanation for the observed distribution of osk orthologs among the Holometabola is that there was a single origin for this gene in a common ancestor of the holometbolan clade , we cannot formally exclude the possibility that the similarity in structure and function between the hymenopteran and dipteran Osk sequences was the result of two lineage specific events of convergent evolution responding to independent instances of selective pressure to establish cytoplasmic inheritance of germline components . However , it seems highly unlikely that the molecular events required to invent a novel gene such as osk would occur in almost identical ways twice in evolution before a different solution is found , let alone the unlikelihood of such a gene being fixed in a population , and then subsequently integrated into a novel regulatory network . However , the invention of a novel factor required for cytoplasmic inheritance of germ plasm components may not be an occurrence unique to the Holometabola . In zebrafish , the bucky ball gene has an osk-like function in generating maternal germ plasm , but is molecularly unrelated to osk , and is only found in vertebrate genomes [11] , [65] . This indicates that there is nothing intrinsic in the primary structure of Osk protein that is required for maternal assembly of germ plasm , and that there are many possible solutions to the problem of generating this substance . Further sampling of metazoan germline establishment strategies will give insight into how common the generation of novel genes is in the process of evolving maternally generated germ plasm . The process of maternal germ plasm assembly must be precisely controlled , and abnormalities in this process result in deep and sometimes spectacular consequences for the embryonic anterior-posterior axis [8] . Based on our results with Nv-bruno and Nv-hrp48 , a common mechanism to spatially regulate osk localization and translation was likely already present at the origin of the Holometabola . This , along with the fact that factors such as Vas and Tud have conserved roles downstream of Osk in Nasonia , indicates that a complex protein interaction network for localized production of germ plasm during oogenenesis existed in a common ancestor of the Holometabola . This raises the question as to when during evolution has this network been assembled , and through which molecular mechanisms . Proteins downstream of Osk , such as Tud , Vas , and Nos , have conserved roles in the specification and function of germ cells throughout the Metazoa , including those without maternal specification of the germline [11] , and therefore are able to function without Osk to generate germ cell characteristics . Similarly , the proteins upstream of Osk , such as Bruno , Hrp48 , and Staufen , are also highly conserved throughout the metazoa , and have conserved functions in mRNA localization and translational control in a variety of cellular contexts outside of the germline . Thus , Osk seems to have been intercalated between two ancient pre-existing regulatory networks . The position of Osk as the nexus between these two networks allows its specific and precisely controlled function in specifying the germline fate . The fact that both the up- and downstream networks were already well established before the evolution of osk indicates that relatively few evolutionary steps may have been required to integrate Osk between them . In addition , since Osk is at least partially derived from a tdrd7/5-like gene , orthologs of which have well described functions in the germline in vertebrates and invertebrates , the ancestral Osk may have been predisposed to interact with other germ plasm components . The localization of osk likely also had an evolutionary antecedent , as the presence of posteriorly localized patterning factors has been detected in some hemimetabolous species , e . g . , [64] , [66] . Since germ cells arise at the posterior pole just after gastrulation in some hemimetabolous species [16] , [67] , [68] , it is possible that factors that predispose posterior nuclei to take germline fate are also localized at the posterior pole in these species . The molecular nature of any such factor , and whether its role is direct or indirect , remains to be determined . Testing the function and regulation of orthologs of genes both up- and downstream of osk in hemimetabolous , and other holometabolous , insect species should give insights into the functioning of the ancestral germline regulatory network , and could provide further clues as to how osk could have been integrated into it .
A BLAST based strategy was used to identify potential osk orthologs in sequenced insect genomes ( see ) . The following databases were searched: for Bombyx mori , Silkworm Genome Assembly at silkdb . org [69]; for Tribolium castaneum , BeetleBase3_NCBI_DB at beetlebase . org [70]; for Nasonia vitripennis , Nasonia Scaffolds Assmebly Nvit_1 . 0 at hymenopteragenome . org/nasonia/ [71]; For Apis melifera , Scaffolds Assembly 2 at hymenopteragenome . org/beebase/; for Acyrthosiphon pisum , genome ( reference only ) at http://www . ncbi . nlm . nih . gov/projects/genome/seq/BlastGen/BlastGen . cgi ? taxid=7029; for Rhodnius prolixus , Harpegnathos saltator and Camponotus floridanus , species-specific Whole-genome shotgun reads ( wgs ) databases were selected at blast . ncbi . nlm . nih . gov; for Culex pipiens Assembly CpipJ1- Johannesburg Strain , Supercontigs at http://cquinquefasciatus . vectorbase . org/Tools/BLAST/ , and for Pediculus humanus , Assembly PhumU1 , Supercontigs - USDA Strain at http://phumanus . vectorbase . org/Tools/BLAST/ . These databases were queried using tblastn with default parameters ( except the E-value cut off was raised to 10 where necessary ) with the following Oskar protein sequences: NP_731295 . 1 ( Drosophila ) , XP_001848641 . 1 ( Culex ) , ADK94458 . 1 ( Nasonia ) , and HM992570 ( Messor ) . To identify osk orthologs among EST sequences , the same query sequences and tblastn parameters were used at blast . ncbi . nlm . nih . gov to search the ( est others ) database . Templates for probe and dsRNA production were generated as in [72] . dsRNA was produced using T7 Megascript kit ( Ambion ) following manufacturers instructions . Fragments used to generate dsRNA and probes were as follows: Nv-osk—bases 161–843 of Genbank accession HM535628 . 1 , Nv-bruno—bases 534–1394 of Genbank accession XM_001605096 . 1 , Nv-vasa—bases 827–1613 of Genbank accession XM_001603906 . 1 , Nv-hrp48—bases 434–1727 of Genbank accession XM_001600216 . 1 , Nv-tudor—bases 6053–6811 of Gnomon model hmm120984 . RNAi experiments were performed as described in [39] . dsRNAs were used at the following concentrations: Nv-osk- 3 . 5 mg/mL , Nv-vasa- 3 mg/mL , Nv-bruno-2 . 5 mg/mL , Nv-hrp48- 1 . 0 mg/mL , Nv-tudor- 2 . 5 mg/mL . Knockdown was confirmed by comparing expression levels of the gene of interest in ovaries of wild-type wasps to those from pRNAi treated wasps . All genes showed clearly reduced levels of expression after their corresponding dsRNA injections , but the degree of knockdown was variable from egg chamber to egg chamber . RACE PCR for Nv-osk was performed using the SMART-RACE kit ( Takara ) according to manufacturer's instructions . The Nasonia Vasa antibody was generated using the custom peptide antibody service of Sigma-Genosys with the peptide CVLRHDTMKPPGERQ as the antigen . It was used at 1∶500 , and detected using anti-rabbit Alexa 555 ( Invitrogen ) at 1∶750 The cross reactive Drosophila Vasa antiserum used in Tribolium and Acanthoscelides was a generous gift from Akira Nakamura [73] . It was used at 1∶1000 and detected as above . in situ hybridization and immunohistochemistry were performed as described in [51] . The ant osk sequence was found in the course of a genome sequencing project ( unpublished ) and cloned from the ant Messor pergandei using the following primers: AntOsk forward ATGGAWGAAACAGTGGCATTRRTMAAAT and AntOsk reverse GGAACCARTCGTAWTCYGTRRTRTACGTT . The cloned 1057 base pair fragment was validated by sequencing , submitted to Genbank with accession # HM992570 and used to generate an antisense Digoxigenin labeled probe for in situ hybridization . Embryos and ovaries of Nasonia and the beetles were collected and fixed as in [72] . Ant embryos and ovaries were prepared and stained for osk mRNA as described in [74] .
|
The establishment of the germline during embryogenesis is a critical milestone for sexually reproducing organisms , but one that is surprisingly labile in evolution . For example , in the fly Drosophila , the germline is set aside early in embryogenesis due to the localized synthesis of the germ plasm at the posterior pole of the oocyte , and the gene oskar is both necessary and sufficient for assembly of the germ plasm . However , oskar orthologs have not been found outside of flies and mosquitos , while the maternal provisioning of germ plasm and the early setting aside of the germline are unique to , but not universal within , the holometabolous insects . In order to understand how the novel mode of germline determination found in Drosophila could have evolved , we have examined this process in the wasp Nasonia . Our results indicate that the phylogenetic origin of the insect mode of maternal germ plasm provision and early establishment of the germline coincided with the origin oskar at the base of the holometabolous insects . Our results further suggest that osk was independently lost in multiple holometabolous insect lineages and that these losses are phylogenetically correlated with changes in germline determination strategies in these species .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"rna",
"transport",
"rna",
"interference",
"evolutionary",
"biology",
"germ",
"cells",
"developmental",
"biology",
"pattern",
"formation",
"gene",
"expression",
"comparative",
"genomics",
"biology",
"molecular",
"biology",
"rna",
"cell",
"biology",
"nucleic",
"acids",
"cellular",
"types",
"genomics",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics",
"cell",
"fate",
"determination",
"evolutionary",
"developmental",
"biology"
] |
2011
|
The Phylogenetic Origin of oskar Coincided with the
Origin of Maternally Provisioned Germ Plasm and Pole Cells at the Base of the
Holometabola
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.